UTAS build 2015

[Jonny Rochester]

715716717

UTAS have made a lot of ambitious design changes and purchases this year. I feel the work I did last year to pull it all together has now led to second year overconfidence, some over stretching in designs, possible sacrifice in reliability etc. Where my emphasis is on using existing techniques, familiar parts, low cost, lowish weight, simplicity, improving everything by 10%, using less of my energy (this being the most important for me), finishing early etc... I feel our team leader is for high cost, new unknown techniques, higher weight (for stiffness), expensive parts $$, wings... (just cause), paddle shift, launch control (sounds cool) and anything carbon/shiny that the other teams are using without necessarily understanding what it is. So right there you can see the tension, added to that a FA that hasn't been able to verify or check our designs... and I had been on the edge of mental breakdown.

We had spent all of first semester (February to June) thinking about our concepts, calculations, keeping a wiki, designing in CAD. Lots of new members, some team leaders not having done FSAE before. Since July we have been building.

While our head team leader does an enormous amount of work, and I stomp on his designs to bring it back to reality, I can say we are meeting a balance and there are alot of positives right now.


Honda CBR600RR - I like this - I am the tuner - engine is running
rear track is narrower
driver pos slightly lower
Chassis is 4130 - added cost - but good I guess
Chassis maybe heavier, stiffer, and we have a CNC alloy plate - I have again been doing the final TIG welding on chassis and wishbones.
we have found more money and more free materials, much more organised with this
we will have new tyres this time!
fresh uprights again, also all new hubs and axles and wheel centers etc. (Staff are quicker on CNC)
using aluminium honeycomb floor bandaged with carbon, steering rack bolted to this (was nervous, did tests...)
the drivers and myself are all going to the gym with a trainer!
Tilton 77's
new rack
weight loss in some parts, other students give you heavier parts... what can you do?
13" wheels
3D print plenum
very very simple Motec M400 setup, no paddle shift or anything yet
undertray, wings...]





So if it works, it will be awesome. If not heavy.

[Jonny Rochester]

718719720

These CAD screanshots are from many months ago. The engine was on a trolley we built in June and I got it running with Motec, laptop, wires etc... Yes, the engine trolley has neon's. A bit of fun.

As an older student it has been very hard to find my place at times. I have not done much detailed CAD design this time, but I have done more calculations, given many ideas some of which are part of this car now.

[mech5496]

Johny, do not make the mistake that all teams do once they feel confident, i.e. do not overcomplicate this!

BTW the car on the CAD looks sexy, but complex and heavy...

[Tyler Jones]

We are running into the same issues on my team this year. Some members want to go from an NA 450 single, 2 element front and rear wing, and manual shifter to a turbo 450 single, 3 element front and rear wing with undertray, paddle shifting and launch control, carbon fiber driveshafts, with a new differential and custom housing.

[Rory Hourihan]

Usually design choices made 'just cause' don't go so well in design. That's the same with most teams I think, everyone wants to over engineer components and tends to forget the trade-off between saving an 1/8 of a lb. and finishing endurance.

[MHybrid Josh]

Usually design choices made 'just cause' don't go so well at comp either. From an aero opinion I am a fan of a highly developed undertray. If I remember correctly the undertray on MR14 (the last undertray i built) was 7-10 lbs. I dont remember the downforce it produced on its own but 60lbs wouldnt be hard to do plus the gains in -lift are much greater than the increase in drag. Ive seen undertrays in CFD with Lift/Drag ratios around 11/1. If your undertray system is well developed that lessens the need for a massive rear wing. Take it from someone who helped build Ann Arbor's "toaster car" as some of us called it in 2014. Manufacturing a big wing like that is almost not worth the effort for the amount of performance you get from it. Keep the aero package simple.


Josh

U of Michigan Ann Arbor
2013 MRacing Aero/Composites Team Member
2014 Michigan Hybrid Racing Front Wing Development Lead
2015 Michigan Hybrid Racing Aero Development Lead

[Rory Hourihan]

Ours hovered around that L/D of 11/1 last year and came in at 8lbs without mounts. Unfortunately, as you say, it takes a lot of time and effort to build aero. So it didn't make it on the car for competition as there wasn't enough time left to validate it :'[

[Z]

Where my emphasis is on using existing techniques, familiar parts, low cost, lowish weight, simplicity, improving everything by 10%, using less of my energy (this being the most important for me), finishing early etc...
...
I feel our team leader is for high cost, new unknown techniques, higher weight (for stiffness), expensive parts $$, wings... (just cause), paddle shift, launch control (sounds cool) and anything carbon/shiny that the other teams are using without necessarily understanding what it is.

Jonny,

You are right, and your Team Leader is wrong!

Sadly, that is the all-too-common reward for the little bit of success you helped UTAS achieve last year.

After last year's comp I wanted to suggest you stick with the "no-plenum" NA 450-single. The fact that the engine lacked a plenum meant that horsepower was restricted to <30 hp, which in turn meant that any pretence of building a mega-horsepower hyper-car would not get in your way. So you could spend plenty of time focussing on all the more important things.

Things like tidying up a car with the same basic concept as Monash's (how many year's have they won FSAE-Oz now?), and also very similar to this year's globally all-conquering GFR. Things like shortening and simplifying the frame (and thus also improving R%!), polishing the steering-geometry, tightening up any toe-compliances, finessing the ergonomics, improving the brakes, +++...

Your 2014 car was good enough to finish just below the half-way mark Overall. Very good effort for a first year Team. IIRC you missed SP and/or Acc, so finishing "the build" a bit earlier might have put you in top ten in 2014. A lot of effort put into improving the many small details of that SAME general concept could well put you into top five this year. Then, in 2016, you could cover it with Monash/GFR-style wings (though I would focus on undertrays, as mentioned above), and do lots of driver training, and UTAS would be looking at podiums.

But I think UTAS should consider themselves a "first year Team" again this year. You are, essentially, starting from scratch... With a much more complicated concept...

So, tell the other members to prepare for a lot of all-nighters, even right in the middle of competition. Perhaps, "Well, there I was, in Monash's workshop, 3 am Sunday, the night before the Enduro event...".

And tell them you are getting too old for all those all-nighters, and they will have to do them... "Yep..., no sleep for two days, six hours to the first Enduro, in a workshop on the other side of Melbourne, and the engine was in a million pieces on the table..."

Z

[Jonny Rochester]

Thanks for your comments Z and others. I think we are all on the same page here, but a team has more than one person. Who ever thought engineering would be managing people?

Our team has done some really good things. In January we wrote a document listing all the improvements to be had in both the 2014 car, and team structure changes, even to how we chair the team meetings and requirements of new members. And we have done most of those things, (except the build is a bit behind and it will be tight...)

Some team stuff, which was new to us:

made new members do a basic application
I helped run a 10 question rules quiz for a few weeks
the core members went to Monash over Easter, "Design to Win". I drove the Monash 2014 car!
student leaders chair the meetings rather than FA
potential drivers now in a gym fitness program
we have a "old" car, so lots of promotional events
wiki webpage to show designs and concepts, not all members use it


Some people thought my questions in the rules quiz where "trick" questions, and some cried when they didn't get 100%. The famous one I'm proud of, "during the tilt test at competition, what G-force does the driver experience when the vehicle is tilted to 60 degrees?" I think the answer is 1G (static situation, despite some discomfort), and all of our team including the suspension guy and the FA got it wrong. Some took it with humor, a girl wrote us a long email wondering why she got it wrong since it clearly says in the rules it's 1.65G, and our whole business team refused to even do the quiz. I laughed, but this and some other pressures meant lots of people effectively quit the team.

But I wanted us to understand accelerations back in February (I don't fully), because our max G-forces dictate all the forces and stress in our designs. I've been using a tyre u of 1.6 for my 'back of envelope' calculations, and I was at first laughed at within our team for it.

[Jay Lawrence]

Hard luck Jonny, I know how frustrating it can be trying to convince people who just want to 'engineer' every tiny piece!

Why do you need a gym program? I ask because endurance is not much of an endurance and if you need a gym program you're probably doing it wrong. Our car had stupid heavy steering for a few years due to some bad design choices, to the point where people were suggesting power steering..... Thankfully they just fixed up the design issue instead.

[Mitchell]

Sounds like the common story, sad to hear it mate. Just keep chipping away at it. The thing that makes me sad about all your changes is your 2014 car needed so very few things improved on and it could be easy top 5. Every team has it's ups and downs, what you guys are achieving is fantastic. It's worth remembering you guys have done more in 18 months than nearly every other team that comes to the Oz comp. Some well established team have done nothing this year, some have made minor changes to last years car, some are still welding their chassis.

It sounds like the groundwork you are laying is the future of a powerful team, just need to work on that decision making. In my view a good team should be finishing design before uni returns in semester 1, building throughout semester 1 and on track in the midyear break. Keep pushing your deadlines (eg: design freeze) back until these dates can make sense, even if you're designing in november. Just be ready to spend $6k+ on tyres.

[Jonny Rochester]

I would say endurance is endurance. Driving the Monash M14 car I worked up a sweat and got a gentle forearm workout, and that is known to be a sorted car. But I still had control of that car so maybe only basic fitness is needed. I can do 70 pushups, and sometimes run 10km during the week, not superfit but basic health is there. Our UTAS 2014 car had heavy steering which didn't show itself on old tyres, but on new tyres we could not drive the car at 100% for all of endurance. The car gripped, we just didn't have the arm strength for the full endurance without making silly mistakes.

So yes, we have tweeked lots of basic things to improve handling and steering effort. But one of our team leaders has got sponsorship with our local gym at uni. Most of us just (me) need to loose 5 or 10kg anyway. The scary thing is, the gym attendance and/or fitness will be a factor in driver selection! as well as time spent actually building the car.

But Uni assignments + build car + fitness + exams, is it possible?

[Jonny Rochester]

Thanks Mitchell, but don't be sad. We'll still beat UQ this year, lol.

It's just with so many changes (and we really are getting some cool materials and parts etc.) I'm just worried about the personal toll due to team arguments, some breakdown in process and team structure etc...

[NathVanVugt]

Nathan here from ECU-R,

I don't fully agree with Mitchell there. ECU pushes hard through design stage but doesn't finalize detailed design until at least a quarter of the way through the year, this allows a good understanding of each system and how they are interacting instead of throwing it all together (not saying you do this Mitchell.) A solid project plan is worth much more in my opinion than getting the car running by mid year, having the whole project broke down into its components and setting milestone achievements, design done, rolling chassis, running car and tracking the project on a weekly basis sets a team up for success.

We also had sponsorship last year from a local fitness company it worked wonders for the two drivers involved (me one of them) who dropped some weight and we both weren't as fatigued as the others by the end of endurance. A lot of emphasis was put onto concentration during the later part of the workout for driving which I think helped me in endurance.

[Z]

Jonny,

The good news is that it sounds like the "Team Building" work is going quite well. (The not-so-good news is "...except the build is a bit behind and it will be tight...". Eeek! Best to panic now, rather than later...)

I liked your "Tilt-Test G-force?" question (yes, answer = 1 G, IMO) . More of that sort of thing at the beginning of the year, plus the current gym/Team-bonding work, can only be good for the Team.
~o0o~

Another thing I would suggest at the beginning of each year is to give each new batch of Initiate Team Members a list of the old car's "Vital Dimensions", but with no numbers given. The Initiates then have to measure the car, and fill in all the numbers on the list.

So start with simple things like overall Length, Width, Height. Then trickier ones like Wheelbase and F&R Tracks. Importantly, the Initiates should also give a TOLERANCE on all the numbers! For example, wheelbase can vary quite a lot with ride-height, steer and static-toe-angles, etc., but by how much? And do you measure it at the outside of the axles, or inside, or middle, or...???

They can then put their tape-measures down and start measuring masses. Total mass of the whole car, and sub-masses of the major-assemblies. Then all the "Centre of Mass" positions, in X, Y, and Z. Wow..., this is starting to get tricky. Then the MoIs, or Radii of Gyrations (very few Teams have a clue about these!). And after all these mass measurements, there are all the important Stiffnesses to measure. Frame-torsional-stiffness, wheel-toe-stiffness, +++. And then ... the engine-numbers...

Having filled in the above list, the now "Official" Team Members will probably still not have a clue what a good Wheelbase number is, or Yaw-MoI, or Frame-torsional-stiffness, etc. But at least they will know WHAT that thing is, and how big their particular one is, compared with the one on the car that keeps winning the competition.

Then, maybe fewer Team Members will go chasing silly numbers that are really NOT NECESSARY (eg. globally all-conquering GFR claim their engine puts out an almighty 51 hp!!!). :)
~o0o~

Other than above, I think Monash's philosophy of "Score Points in ALL Events" is a good one to push, from now til December.

Z

[Jonny Rochester]

Good suggestions there Z. I did do alot of thinking and talking along those lines in first semester. Unfortunately not alot of our old car got measured (I wanted to measure rear toe/camber compliance and steering link forces, and maybe chassis stiffness). Our old car has largely been left for promotion and fund raising, as a static display.

But funny thing is, we need to drive the car for uni open day next week. So we are giving the old car a new clutch and a few bits. In pulling the exhaust wrap off, only now we find the aluminium muffler had melted. So some learning and experience will take place I'm sure.

[Mitchell]

ECU pushes hard through design stage but doesn't finalize detailed design until at least a quarter of the way through the year, this allows a good understanding of each system and how they are interacting instead of throwing it all together

Why not start design in October and finalise in February? Your end point is going to be determined by the starting point. It is much easier to get a group of people putting in 200-300 hours a week during holidays (which start ~mid november...) than during uni. For a new (maybe any?) team 3 months of driving/testing time is going to be significantly more valuable than 3 months of analysis.

[tromoly]

For a new (maybe any?) team 3 months of driving/testing time is going to be significantly more valuable than 3 months of analysis.

This. In my opinion the easiest way to make a car consistently "faster" is by getting the driver(s) comfortable near the car's limit.

Side note on the gym discussion, in the bicycle world there's a tendency for some bike riders who are a bit overweight to walk into shops and spend obscene amounts of money on parts that in total only shave a few grams to maybe a pound off their bike, when asked they usually say they're "trying to cut weight for the season" while ignoring their own additional weight that could be lost.... So the next time systems are broken down on where weight can be removed, it might be worth adding a column for the drivers as well.

[NathVanVugt]

Why not start design in October and finalise in February? Your end point is going to be determined by the starting point. It is much easier to get a group of people putting in 200-300 hours a week during holidays (which start ~mid november...) than during uni. For a new (maybe any?) team 3 months of driving/testing time is going to be significantly more valuable than 3 months of analysis.

Fair point Mitchell, but the competition isn't all about driving/testing time as I am sure you are well aware. If I was a new team I would take my time throughout the year making sure my detailed design is as good as it can be, running simulations and physical tests as often as possible to make sure the team is prepped for design, cost and dare I say it presentation. If this means the team looses a few weeks driving and testing time so be it, having the car planned out to be finished three quarters throughout the year with the team prepared to maximize points in statics and just under 3 months driving time would be the place I would want the team to be in.

Of course this is my opinion and I haven't been in a team that is just starting out but I don't think the plan would deviate that much from the way some of the top teams run.

[Menisk]

I think the difference in opinion is also reflective of different approaches. Some people very much like to consider everything carefully and make sure they get it right once and other people prefer more of an educated guess and check approach. I know we're very much a fan of the educated guess and check approach as there are some things that are difficult enough to analyse that you're bound to overlook something. The earlier you can get bit by that oversight the earlier you can get to fixing it. There are some things we can't analyse because we don't have enough data to do so. For example we're still waiting on tyre data for the 7.5" rears we're choosing to run. We still aim to do as much analysis as we can and have a thorough understanding of what we're building before we start manufacturing, but we'd rather stop at 85% understanding and gain the last 10-15% understanding through physical testing.

Ultimately the guess and check approach allows us to get a ballpark car on track earlier and use the fantastic feedback of driving it to make the required modifications to make it fast quickly. We also get the added bonus of driver training, something a team without a lot of FSAE driving experience will benefit from massively. We didn't get much drive time last year and as a result our drivers weren't able to comfortably push the car and get the most out of it at competition.

[NathVanVugt]

For a new team starting out I am not sure that the risk of getting out that early would benefit the team. It also depends on the level of team just starting out too. Maybe in your case that 10-15% is the physical testing of tyres, but this 10-15% could be the oiling system or electrical faults which if done in a systematic approach at first could have stopped a catastrophic failure or days/weeks of fault finding. Not everyone has a big team with experienced team members, space out the build and hit milestones!
We obviously have different strategies to tackle the same problem and at the end of the day they could both have the same result, their is no denying that getting a car out early will improve driver training (if you don't have a previous car to use) but making the car as simple as possible, low parts count and ease of manufacturing would be my advice for UTAS, don't over complicate the already complex problem you are trying to solve.

[Jonny Rochester]

725

Most of our car is being made by students, as it should be. Students do all of the welding, including chassis, wishbones, custom radiators, fuel tank. All of the CNC aluminium machining is done in house by qualified staff, but student designs of course. The engine plenum is one thing we outsourced as we didn't know of a good enough 3D printer at our uni. It is nylon. We have been told to use a solution of super-glue and water to seal it. This part is a big step for our team as the "plenum" on our single cylinder last year didn't actually exist! AT-Power throttle will be bolted on top.

[Menisk]

Let us know how the sealing method works for you Jonny. We've tried water based varnish in a 4:1 mixture with water and we're unsure about how well that worked (the plenum leaks from the flanges). Last year we sealed ours by painting it, that one worked great.

[Tyler Jones]

Let us know how the sealing method works for you Jonny. We've tried water based varnish in a 4:1 mixture with water and we're unsure about how well that worked (the plenum leaks from the flanges). Last year we sealed ours by painting it, that one worked great.

Our team has used a 3M 2-part epoxy for the past two years and it has worked great. I would advise against the superglue and water mixture, superglue reacts with water.

[Adman]

Ah! The (annual?) Utas build thread, where Jon makes us appear to be in total dissarray and have no chance of making competition! ;)

I should introduce myself, I'm Adam, Head Engineer of Utas this year. I'm definitely am ambitious leader, by no means perfect, striving to make some solid changes this year to put us closer (perhaps beyond) to 10th place this year.

Where to start? Firstly, it is true we are making some big leaps this year, some into the unknown, some to greatly improve what we had last year. It is worth taking what Jon says with a pinch of salt... as it is often biased towards his way of thinking! I guess I'll just run through some of the points that have been raised and offer my 2 cents on them.

Head engineer who is doing things 'just cause'? That's a little unfair. There are certainly some things I threw out as ideas at the start of the year that were pretty adventurous and perhaps unnecessary. Jon has made it appear like I am trying to break the team with ridiculous ideas. The reality is it is mainly Jon and myself who question designs, check their association with surrounding components and oversee the majority of fabrication. There are many, many things that are so much better for us this year. We have some very keen and dedicated team leaders and a team size nearly double that of last year. Our sponsorship is nearly triple that of last year (so yes, we will have tires that don't resemble pie pastry!). This time last year, we had car 'on wheels', yet we do not this year. However I'm more confident we will finish in time for adequate testing this year as numerous parts of the car last year did not actually function at this point (pedalbox broke, steering column didn't have a brace, engine wasn't running).

We still have problems with people ignoring fine details - poor drawings, attention to bolt sizes and tolerances, and a 'she'll be right' attitude towards getting things made on time, at all, or actually working properly!

Unfortunately I've had to take on a huge workload this year; Jonny was originally ergonomics leader but our keen-as-a-bean electronics leader decided to pursue motorbike racing. So I was then chassis and ergo leader (as Jon took over electronics), as well as HE. Then our driveline leader graduated early and left the team, with all the hubs and driveline "designed", FEA'd, but poorly thought through in the details. It was a mess, and a massive redesign I had to do which I didn't have time for. Regardless, we're still progressing. Our suspension leader works behind closed doors for the majority of the time, which makes it hard to communicate with and check details of design.

We are certainly a rookie team still. No doubt about it, but I think we are still operating without our capabilities. We have gone for an aluminium bulkhead to house drivetrain this year, something that I have been an advocate for, but it is a conservative design weighing in it 3.2kg (Monash weighs sub 2 for reference), safety factors of 4 in most cases using 7075 aluminium. Aero is being implemented as part of a final year reserach project this year and as such has had plenty of first principles, CFD and hopefully soon, some dynamic testing associated with it. Undertray is done. Bodywork molds in progress, wings are a bit of a worry (see 'lack of fine details' comment above). Engine already running this year (thanks Jon!) whereas last year I think we first got our engine running in early November!

Myself and the team leaders want to arrive at competition with a car that is worthy of the time we have spent working the longer hours that we are all too familiar with as FSAE members. Jon is a great sanity check, toning back some of our more audacious ideas, and has so much more experience than any of us in actual mechanic-ing and fabrication. Him and I have frequent 'heated discussions' but we get on well enough :) I will leave you with a picture of our 2014 car, that we have affectionately named 'Dugong'. I think you will see why. Hopefully 2015 will be a lot sleeker and faster!

727

[Z]

I'm definitely am ambitious leader...
... striving to make some solid changes this year to put us closer ... to 10th place this year.

Adam,

The UTAS 2014 car, in the condition it was in on the Sunday of last year's comp (ie. running at last!), would be more than good enough to come 10th this year. If you do not achieve significantly better than 10th this year, then all the work the Team has done will be wasted.

Overall, it appears to me that UTAS's "Design Management" has headed in completely the wrong direction this year. I can only assume that you did NOT read Geoff's "Reasoning Your Way Through the Design Process" thread? Shame, because last year's effort was a good start.

Z

(BTW, the 2015 frame design is an abomination (no better than last year's), and the machined rear-bulkhead is unjustifiable nonsense (... more wasted time and money).)

[CWA]

Adam,

You come across well. You seem to have a good team-lead head on you, and seem to uphold a very professional demeanor that I didn't experience too much myself whilst competing. That said, I'm sure you are conscious of the dangers of the 'all talk no walk' approach.

Ignoring technical details of the project so far, I am curious to know how effectively you think the sub-teams / CoCs are being managed; how your entire workforce resource is being utilised. I only mention this because of the way you described the progress of your individual CoCs (you mentioned ergonomics, driveline, electronics, and suspension CoCs), and the emphasis you placed on ownership of this progress lying with specific people.

Do you think a change in the way responsibility is shared between your ("larger than last year's") team might minimize the impact of having CoC leaders becoming distracted / change roles mid-project? I wonder if perhaps you think the CoC leaders themselves currently have too much responsibility, responsibility which could potentially be delegated down to other team members who are just as capable?

Regards,

Chris

[Adman]

Z,

What you saw of the Utas14 (endurance and the go kart testing day) was a snapshot of glory! Since then, in every testing session or display, the car has broken without fail. We managed 14th place mainly as a result of many of the teams failing to complete endurance. While that is obviously a common situation, if other cars (Canterbury, Adelaide, Wollongong) had done better, we would have been significantly further down the order.

An abomination? Interesting feedback, perhaps you can tell me why? It's significantly stronger, better packaged and slightly lighter than last year, so why is it so poor? Let's hope Utas15 does indeed do better than 10th then, otherwise the failure is on me! I have read Geoff's reasoning, but there are some fundamental performance issues we have had to fix that make us appear to be going down a completely different design process. For instance on the 2014 car,


suspension geometry was based off 2002 car, not ideal, but saved us time. Not all suspension loads on nodes either.
many components reused from 2002 (hubs, shafts, callipers) and did not allow us much design freedom
engine... oh that engine. it is wrecked. very poorly tuned, breaks every time. had to go
steering? by end of endurance we were hanging off the wheel trying to turn it. needed new steering geometry and design
aero created lift... not downforce. poor radiator flow. intake behind driver (less flow)


So you take all that, attempt to improve it, and try to fit it on last year's chassis? It was simply not worth doing. Hence the redesign we have undertaken this year. Will it be an abomination? I'll do my best to prove it will not be...

Chris,

Thanks for you comments! We run our team in sub-teams as normal (chassis, ergonomics, suspension, drivetrain, engine (split also into mechanical and electrical for wiring) and aero), plus a head engineer and team leader. Each sub team has a leader that oversees the design of that part of the car, communicates with other leaders to ensure coexistance and fitment of parts. Sub teams are usually between 3 and 6 people depending on work load of that car section.

The idea in principle should work well, with a range of students across their degree in each sub team. However, last year the majority of work was completed by leaders alone who have since graduated (something we've tried to rectify this year by getting more students involved in design). As a result, there are very few students on this year's team who understand the design process, so it is a challenge to give components to these fresh students. Generally, leaders give a design brief which the younger students should follow and work with their leader to come up with the design.

When our electronics leader left, it was not such a problem. This was early on, not much had been done, and Jon went to electronics and I took over ergo. I had lead ergo last year anyway, so it wasn't too much of a strain to know where to improve on last year.

Driveline was different; our leader was doing his final year project on the redesign of the hubs and spool assembly, so his team hadn't had any involvement with that design. When he left, it was 'in CAD' and looked okay, but was far from complete for a large number of small details. There was nobody who really understood the system other than the leaders and myself, and we had very little time to fix it, so I took that on to push it through to production. For instance, he had designed brake discs, but had not investigated how to make them locally or if we even had the materials! It turned out to be impossible to do, and so I had to redesign them with production discs we could buy.

In short... we have a hierarchy that works okay, but when our driveline leader left it was a special case where the responsibility was solely on one person which meant big problems for the team.

Adam.

[Kevin Hayward]

Adam,

Out of curiosity why did you choose to form subgroups based on subsystems rather than on activities?

If you base it on subsystems then you are very exposed whern a team member walks. You have the disadvantage of thinking at the subsystem level rather than holistically, as well as almost dictating the final design by the different subgroups.

Furthermore you will tend to find that subgroups based on subsystems require expert knowledge which most students lack. Instead engineers tend to show performance in a particular activity. For example you will easily be able to identify which students are good designers / good manufacturers / good testers / good management. When split up into subsystems you need the engineers to be good at all activities.

Instead you could split your team into something like the following:

Management
Design team
Composite team
Machining team
Assembly team
Test Team
Statics team

This means people can be placed in activities that they are good at. A single team member may be involved in maybe 2 or 3 teams at most. This way instead of losing your single drivetrain expert you just lost a designer, or a machinist. One of a small group that already will meet the new work load.

My general feeling is that the larger FSAE teams can cope with being split up into subgroups because there are multiple people in each area, however for small teams the risk of losing your only team member with any knowledge in a particular area is too great.

You may also want to read some biographies of designers in motorsport. Tony Southgate's is a pretty good read, as is Ron Tauranac's. You will find that these designers could design just about any part on the car, as you will find machinists that could make any part on the car.

We attempt to model this way of working at ECU. The Uni has quite a small team, being a very small engineering program. We tend to keep people involved for quite a long time. For example our technical director this year is primarily a designer / machinist. In his time in the team he has designed aero packages, suspension systems, chassis, manufactured wings, assembled engines, machined parts for a lot of systems. There wouldn't be a single part on the car that he couldn't design. We tend to start students out in assembly / manufacture / testing and slowly move them into significant design roles.

Designers move from one part to the next as they finish up, as do machinists (and so on).

When we lose a member (and we do) or someone fails to deliver it isn't hard to pass the project onto another designer / machinist.

There is almost no possible way that we could run like Monash or the other big uni's, we simply do not have the number of students. I doubt you should be modelling your structure off those teams as well. Small companies do not have the luxury of hyper-specialization, but if set up right they can be much more responsive to change.

Kev

[Adman]

That's really interesting Kevin. I think the problem we have faced is that last year we (obviously) had to create a car from scratch, and we pretty much done the same again this year as we felt there were very few components we actually wanted on the 2015 model. As such, we have had a massive design stage, and subsequently a massive fabrication stage. Pretty much no manufacture occured before June, and no design has occurred post July (except for some emergency redesigns). This would mean for us that the design team doesn't have a whole lot to do later in the year, and the manufacturing team would stand around in the early stages, so it makes a bit more sense to have students design and component and then follow through with its fabrication.

Composites and machining teams would encompass the fabrication stage, and as I said above, we pretty much try and get everyone involved in fabrication. Students who understand it better use more advanced machinery and work on more critical components, but some will start on the most basic components and work their way up.

We have not done much testing of the old car because we've not had the funding to fix it until 2 weeks ago, but we should be looking at getting a team of students who can run a vehicle and collect the associated data with it for later this year in preparation for the new car.

A statics team? I assume you mean design/business presentation/cost report? This is an excellent idea... I'm trying to get momentum started on the CR at the moment and it's a struggle, but perhaps a more dedicated team to do that would work better than trying to spread the load across everyone.

Next year we do not plan on a full new car as we feel we will have reached a reasonable level of performance and a maximum complexity that we are capable of at this stage. Perhaps then would be a good time to initiate a different team split, as parts would be designed for more of the year, fabrication started earlier, more testing done, and overall less stress!

[Z]

An abomination? Interesting feedback, perhaps you can tell me why?

Adam,

Much as I would like to point out all the errors in your latest frame, I won't, because I know it will be a waste of time. I know this, because I clearly spelled out ALL THE SAME ERRORS in last year's UTAS thread! You can lead a horse to water...

At least last year's frame didn't have that nonsensical machined aluminium bulkhead. In 7075! Do you realise that all the CADing and FEAing you did to turn that ma$$ive block of expen$ive alloy into a 3.2 kg wet noodle, could be replaced by 3.2 metres (!) of FSAE-standard 25 x 1.6 mm mild-steel tube, of the SAME MASS? And said steel-tube-bulkhead would be much cheaper and quicker to build, and very much stronger and stiffer...
~o0o~

A clue to your poorly reasoned design approach this year is this.


Next year we do not plan on a full new car as we feel we will have reached a reasonable level of performance and a maximum complexity that we are capable of at this stage.

Why do you want "maximum complexity"? Why???

(Edit: In case I am not clear here, why not aim for "maximum simplicity"?)

The whole point of Geoff's "Reasoning..." thread was to stress to you students that you should be very carefully counting the points!

Namely, how many competition points might you gain by using [insert any "FSAE-thing", such as "machined 7075 bulkhead"], and how much will that thing cost you in time and money? This process is known as "Cost-Benefit Analysis", or "Return-on-Investment".

You might want to try some examples:

Specifically, what is your estimated "competition-points-gain vs cost", or "RoI", from that machined bulkhead?

More generally, what is your estimated RoI from your addition of "maximum complexity"?
~o0o~

To repeat (IMO):

A reliably running version of last year's car would be good enough for 10th place this year.

A better detailed version of last year's car, with lots of driver training, could be a top-five contender this year.

Z

[Jonny Rochester]

Z,
I am with you on the rear bulkhead. I had personally voted for a steel welded bulkhead as part of the chassis, (maybe lighter, stronger, simpler, cheaper, quicker to make etc. as you say) and I would have used square tubes like the Curtin car. (Secretly like Curtin while I know it's not following your philosophy Z). But the decision was made at a higher level, and I'm happy to work with that. But this brings me to another sore point, and that is I didn't agree with the hand calcs done on the machined bulkhead, and so I couldn't agree with the FEA either, but it was made anyway. And I/we felt the faculty adviser would/should advise us on mechanical design, but he couldn't. Now I'm told another bulkhead will be made as a spare, just in case, but Exactly the same! I suggest MkII bulkhead be different in some way after reasonable hand calcs.

...but this is airing dirty laundry, and these are current issues still in play. I have been hauled in the the head of school office once already for my remarks!

[jd74914]

Perhaps this isn't germane to the current argument, but I've often thought a well-contrived rear bulkhead [coupled with good engine and chassis model] would greatly speed up fabrication due to the lack of welding and increase precision due to the inherent jigging. I've always wanted to run a cast one with minimal post-machining, but that's a different story.

[Mitchell]

I have always thought the best rear bulkhead is one that doesn't exist. It's lighter, cheaper, easier to make and the accuracy doesn't even matter.

[jd74914]

Yes, I should have added the caveat that the bulkhead is needed (for toe base, etc.).

[Z]

Jonny,

Yes, RHS or SHS works well for flat frames (I don't like them for more general 3-D spaceframes, because the nodes look messy).

So, working this out on a post-it note, a removeable (*) rear bulkhead of about W = 0.6 metre x H = 0.4 m, made from SHS = 25 mm x 1.6 thk (= ~1.25 kg/m), plus some diagonal bracing from smaller sections, plus the numerous inserts to take bolts, +++, comes out at about ... 3.2 kg!

(* Edit: Only "removeable" if that is needed for, say, engine removal. Better is "welded-in", or "not there at all" as below.)

This can be jigged very easily on a flat table (just clamp down...). And the precision can be exactly the same as the machined aluminiun part because all the important holes/surfaces can be inserts of round bar that are welded in first, then drilled/milled last (after a stress relief, if you want super precision).

As for the co$t of that $lab of 7075, mo$t of which gets turned into $warf!!! ($1,000+???)
~o0o~


I have always thought the best rear bulkhead is one that doesn't exist. It's lighter, cheaper, easier to make and the accuracy doesn't even matter.

YES, yes, yes, and yes!!!

And best of all, the "part that is not there" is extremely RELIABLE! It never, ever breaks. Never... :)

Z

[Adman]

Z

Thanks for your comments. I'm well aware of the numerous issues with the 2014 frame, and I've done my best to try and rectify these on the new abomination. Things like suspension load paths not placed/or leading to nodes, poor triangulation. I have also talked to several other teams and FAs on their ideas, plus done my own research and ANSYS testing as to which are the most important tubes to put in place for the loads we will see (and of course, the so desired torsional rigidity). As I've said, this chassis is slightly lighter that last, offers nearly 4 times the torsional rigidity, and is a better packaged unit to encompass the components we've placed within it.

Rear bulkhead; using a 4 cylinder made it difficult to package effectively. When trying to come up with an assembly that worked effectively for suspension pickups, jacking bar, driveline attachment, engine bracing and removal, we looked at a number of options. I investigated steel tubing as an option, but deemed it heavier overall once the large number of tabs and additional welding required to package the aforementioned assemblies. It is not in level 3 of Geoff's design reasoning that it is important for parts to interact effectively, and to look at design at a 'whole vehicle' level, rather than just "how light can I make the chassis"? There are pros and cons I guess; we landed on a removable bulkhead to allow for a bottom out engine install among the other reasons above.

I think you've taken my complexity comment out of context. I simply meant that Utas does not have the capabilities to extend into mass carbon fibre designs, hardcore aluminium CNC, 3d printing titanium, etc. The 2014 car was a poor representation of what our team can potentially do, this year will be more at the level we expect of a Utas FSAE car, and future years can refine what we have done in the manner you would have liked us to do this year. We have tried a few interesting things this year, tried to be a little different but also keeping on what we know we can do (basic steel spaceframe, basic aluminium CNC).

To repeat, I do not mean that we have aimed for 'maximum complexity'. I mean that we will have reached a level of complexity that this team can comfortably work at. To be working on tight safety factors, composite chassis', turbocharging, custom engine blocks, carbon fibre components everyone.etc. is simply beyond us.

[Menisk]

I'm well aware of the numerous issues with the 2014 frame, and I've done my best to try and rectify these on the new abomination. Things like suspension load paths not placed/or leading to nodes, poor triangulation. I have also talked to several other teams and FAs on their ideas, plus done my own research and ANSYS testing as to which are the most important tubes to put in place for the loads we will see (and of course, the so desired torsional rigidity). As I've said, this chassis is slightly lighter that last, offers nearly 4 times the torsional rigidity, and is a better packaged unit to encompass the components we've placed within it.

Do you understand why you might want torsional rigidity in your chassis? From that understanding did you have a torsional rigidity value that you aimed for?


Rear bulkhead; using a 4 cylinder made it difficult to package effectively. When trying to come up with an assembly that worked effectively for suspension pickups, jacking bar, driveline attachment, engine bracing and removal, we looked at a number of options. I investigated steel tubing as an option, but deemed it heavier overall once the large number of tabs and additional welding required to package the aforementioned assemblies. It is not in level 3 of Geoff's design reasoning that it is important for parts to interact effectively, and to look at design at a 'whole vehicle' level, rather than just "how light can I make the chassis"? There are pros and cons I guess; we landed on a removable bulkhead to allow for a bottom out engine install among the other reasons above.

Your packaging problems don't come from having a 4 cylinder they come from choosing a double a-arm rear suspension. If you want things to package nicely have a think about what motion you're trying to constrain and different linkage systems that will constrain that motion. There are plenty of different suspension configurations out there that package far nicer in the rear of an FSAE car than double a-arms (hint: Have a look at 4 link de dion beam axles). In terms of kinematics do you understand the advantages and disadvantages of double a-arms and was this a consideration in your choice of rear suspension configuration? We have a car with a 4 cylinder, no rear bulkhead and engine installation is probably the easiest of all of our cars. We literally drop the rear spaceframe onto the engine and put the bolts in. 5 minutes tops.

[Z]

Adam,


I think you've taken my complexity comment out of context.
...
I do not mean that we have aimed for 'maximum complexity'. I mean that we will have reached a level of complexity that this team can comfortably work at.

Your attitude to "complexity" is quite clear. You are striving for the MOST complexity that "...this team can comfortably work at.".

This year it is machined aluminium bulkheads. In future years it will "... extend into mass carbon fibre designs, hardcore aluminium CNC, 3d printing titanium, etc."

But nowhere have you given a points-based justification supporting this irrational ideology that more complexity = better.

Never mind. That sort of thinking is very common in FSAE. Witness all the cars that are incapable of driving 30 km at an average speed of 50 kph!
~o0o~

I guess there is no point going further into these big-picture issues here, but I am interested in some of the smaller details.

You said your new frame is "...nearly 4 times the torsional rigidity" of last year's frame. So, some questions.

1. What is the measured torsional stiffness of last year's frame?

2. How did you measure it?

3. Is your 2015 frame finished (less than 90 days to go!), and if not, then where does the "4x" estimate come from?

Note that many students' FEA estimates of torsional stiffness are "imaginary" numbers, mainly because of the unrealistic constraints used. You might shed some light on this by sharing your FEA numbers and the real measured numbers.

Z

[Adman]

Thanks for you comments guys,


Do you understand why you might want torsional rigidity in your chassis? From that understanding did you have a torsional rigidity value that you aimed for?

It is my understanding (a summary of) that bump absorbtion should be made through the suspension dampers and ARB where it can be controlled, rather than through a flexing chassis where you have no refinement about how much roll stiffness you have through cornering.

Based on conversations with Scott Wordley, reading papers and forums, I've aimed for a 2.8kN/deg in ANSYS, expecting that to drop to 2.2 for the real bench test. Last year's was about 900N/deg by comparison (ANSYS simulation).


You said your new frame is "...nearly 4 times the torsional rigidity" of last year's frame. So, some questions.

1. What is the measured torsional stiffness of last year's frame?

2. How did you measure it?

3. Is your 2015 frame finished (less than 90 days to go!), and if not, then where does the "4x" estimate come from?


It would have been ideal to take the old car apart and bench test it for torsional stiffness, comparing that to the ANSYS FEA I've done on it. While last year's head engineer did his own FEA on the chassis, his constraints were somewhat kinder than mine and gave a disproportionately higher TS than what it should have been. I will post FEA figures later, but as far as values go, last year's chassis got 900Nm/deg, this year was about 2.65kNm/deg (apologies, I had thought it was 4 times, only 3). The majority of last year's poor results came from the rear enclosure of the driveline; the system was poorly triangulated and so the suspension wishbones were applied to the 4 sides of an open cube.

Chassis is 'done' but still plenty of attachment points to be welded on. Closeout panels will be more structural this year to aid in structural performance, but was not included the in ANSYS simulations.


There are plenty of different suspension configurations out there that package far nicer in the rear of an FSAE car than double a-arms (hint: Have a look at 4 link de dion beam axles). In terms of kinematics do you understand the advantages and disadvantages of double a-arms and was this a consideration in your choice of rear suspension configuration?

Haha, I would love to try a de dion beam axle. Yours (UQ?) looks fantastic and simple, and I loved ECUs. I guess we didn't consider changing from double wishbones as it's what we did last year, and know (to an extent) what we're doing. But can you imagine the response I/we would've got if we'd tried to change yet more things :P Perhaps an option for next year to improve reliability and simplicity!

[Menisk]

It is my understanding (a summary of) that bump absorbtion should be made through the suspension dampers and ARB where it can be controlled, rather than through a flexing chassis where you have no refinement about how much roll stiffness you have through cornering.

Based on conversations with Scott Wordley, reading papers and forums, I've aimed for a 2.8kN/deg in ANSYS, expecting that to drop to 2.2 for the real bench test. Last year's was about 900N/deg by comparison (ANSYS simulation).


The root of this decision should be based on how you want to be able tune the distribution of weight transfer in your car. I'm not sure if you've had a read of RCVD or another book on vehicle and tyre dynamics, but the long and short of it is that you'll likely want to be able to tune the amount of load transfer happening at each end of the car in order to balance understeer and oversteer by taking advantage of tyre load sensitivity. On a regular 4 corner spring car you do this by changing your front and rear roll stiffnesses, (the lateral load will balance towards the stiffer path).

Think about your car as 3 torsional springs in series, front roll stiffness, chassis stiffness and rear roll stiffness. What happens to your ability to tune that distribution of load transfer with your roll stiffnesses when your middle torsional spring (the chassis) is almost the same stiffness. What happens when it's an order of a magnitude more stiff, what happens when it's 2 orders of a magnitude stiffer?

When you have a 50/50 weight distribution car with the same tyres on all corners, (like almost every FSAE car) how much ability to change your lateral load transfer distribution do you actually need? Our previous car was 52% rearward weight biased and we normally ran an LLTD somewhere in the order of 55%, it really didn't need much.

While numbers from other teams can be nice to see if your calcs are at least in the sensible ball park you definitely need to start from scratch to see how people have been coming up with those numbers and whether their numbers are going to work for your concept or if you need more/less than they do.


It would have been ideal to take the old car apart and bench test it for torsional stiffness, comparing that to the ANSYS FEA I've done on it. While last year's head engineer did his own FEA on the chassis, his constraints were somewhat kinder than mine and gave a disproportionately higher TS than what it should have been. I will post FEA figures later, but as far as values go, last year's chassis got 900N/deg, this year was about 2.65kN/deg (apologies, I had thought it was 4 times, only 3). The majority of last year's poor results came from the rear enclosure of the driveline; the system was poorly triangulated and so the suspension wishbones were applied to the 4 sides of an open cube.

When applying your constraints, where is the roll force being reacted? It's not at your wishbone mounting points. It's at your shock mounting points. Wishbone mounting points react lateral and longitudinal loads.

[Z]

Adam,

Torsional-Stiffness (axle to axle of a car) is measured in kN.metres/deg (or radian).

More importantly, you do NOT need to "...take the old car apart and bench test it". An FSAE car can have its TS measured very easily, with reasonable accuracy, with a single bathroom-scale, bubble-level, steel-ruler, and some other off-the-shelf stuff. A heavy Team member sitting in the seat, or two or three standing there, gives multiple load conditions.

If no measurement of real frame TS, then all your FEA is meaningless.
~o0o~

As for a single-cylinder car with De-Dion and 10"s, well, ... I look forward to seeing Auckland in eighty-odd days! :)

Z

[Jonny Rochester]

At the beginning of the year, I was looking at driver position, and so wanted to know CoG of the 2014 car. We put Adam in the car and tilted it up to balance point, and from this I calculated/estimated the CoG was ~345mm (much higher than stated in our design report). So we measured something! I was also looking at steering effort, so I measured static steering torque at the steering wheel, with a torque wrench, ~19Nm. Two things measured! Plus weight balance which was easy, front heavy. And I did designs/concepts to change these numbers in the new car.

It would be good if the chassis guy tried to measure torsional stiffness, and I suggested the suspension guy get a huge bar and measure camber and toe compliance etc, and I suggested to tape an iPhone onto the old car to get accelerations, but all these things are just too hard to organize. I am often on my own in making these suggestions so it saps my energy just talking about it.

For chassis torsional stiffness, I havn't invested too much time thinking about it (I'm not chassis leader obviously), but I think that is important for avoiding understeer and getting fast transitions between left and right turns. Our 2014 car did understeer and was slow to transition between oversteer, then understeer, then a delay before the car steered. In some situations this is hardly noticeable, but in others I did notice it alot and it can be improved. I think this was due to having a basic chassis with minimum number of tubes (yes it was triangulated, yes it can be improved bla bla...) , just a little light for our level, and it did crack several times.

This year we have a 12mm aluminium honeycomb floor which will be bandaged to the chassis with carbon, and this will probably add some stiffness (and weight?) to our chassis. I do believe Adams chassis will be stiffer this year, mostly from a higher number of tubes placed within the wheelbase, the harness bar is more integrated, and a bit from using 4130.

As for suspension points being on a node, our 2014 car was simple and not too bad in this regard. UTAS 2015 car maybe worse, if suspension on a node was your thing.

As for beam axle and de dion, I don't think we have enough leaders that stand back and think conceptually in that way. It takes alot of prior knowledge to know of different possible setups and then be able to choose one. And so I think it is OK to stick with what is normal and known to be made to work.

[Z]

Jonny,

MEASURING FRAME TORSIONAL STIFFNESS.
==============================

This is how easy it is.

1. Gather together required tools of bathroom-scale, bubble-level (say, 1.2 m long), steel-ruler, and miscellaneous junk as noted below.

2. Establish which end of your racecar is heavier, and support those two wheels on blocks of wood of similar (vertical) thickness to the bathroom-scale. Then the bathroom-scale goes under one of the wheels at light end of car. So for front-heavy UTAS-14, the blocks of wood go under front-wheels and bathroom-scale under one rear-wheel.

3. Remove bodywork if necessary, and locate two horizontal surfaces of the frame, one at front of car, and other at rear of car, that will have their relative "twist" measured. Typically, these might be the tops of Front-Bulkhead and Rearmost-Bulkhead, although you can also measure the twist at intermediate locations. It may be helpful to clamp a 1.2 m long x 25 mm SHS to these horizontal surfaces, for more accurate measuring. But just measuring off the top surfaces of the bulkheads is good enough to start.

4. Place bubble-level on the top surfaces of the bulkheads, or on the SHS clamped to them. Put the steel-ruler, in a vertical orientation, on top of the lower end of the frame or SHS, clamp the bubble-level to the steel-ruler with your fingers, and adjust the bubble-level height so that it becomes horizontal (ie. bubble in centre of lines, as seen with one eye closed, no parallax errors, etc.).

Shout-out the reading on the steel-ruler, namely of the height of bubble-level above that end of frame, to your assistant who writes down that number as the first data point. (Or do this yourself, while muttering profanities about useless young people...). Also write down the bathroom-scale reading, and other relevant stuff, as noted below.

Note here that if you are using a 1.2 metre long bubble-level and SHS-clamped-to-frame, then putting a mark 1,146 mm from the end of the bubble-level, and measuring the vertical gap there, makes conversion to degrees very easy. Namely, 20 mm gap = 1 degree away from horizontal. If you only have a shorter bubble-level, say 2ft/0.6m, then 10 mm at 573 mm = 1 degree. All this from 2 x Pi radians = 360 degrees.

5. To get more data points get some heavy Team members (= the "miscellaneous junk" noted above) to sit or stand in the seat area as "centrally" as possible. This increases the reading on the bathroom-scale, and increases the amount of frame-twist. Then swap the bathroom-scale to under the other wheel of that axle, so all the twists go the other way. Write down all important numbers.

6. CALCULATIONS - The bathroom-scale indicates a vertical-upward force acting on the underside of that wheel (ie. from ground, to car). This force acts at a lateral distance of T/2 (= "half-track") from the centreline of the car. (This distance depends on the position of "centre-of-pressure" of the tyreprint, which, to a first approximation, is ... the middle of the tread. You can get more accuracy for this dimension by putting something "narrow" under the tyre to simulate a point load.)

By standard Statics methods, this "single upwards Force Fw acting at centre-of-wheelprint" can be replaced by an upwards Force Fc, of same magnitude as Fw, but acting on the car-centreline, together with a pure Couple = Fw x T/2, acting in the lateral-vertical plane (ie. drawn as a longitudinal vector).

So, for each such calculated Couple (= bathroom-scale reading x half-track measurement) you have a corresponding measured twist of the frame (= the difference of the bubble-level-measured front and rear slopes (Edit: see clarification next post)). Draw a graph with "Applied-Couple (kN.metres)" on vertical-axis, and "Frame-Twist (degrees)" on horizontal-axis. Insert your measured points, and draw a "line-of-best-fit". The slope of this line is your frame torsional stiffness (in kN.m/degree).

7. ERROR ESTIMATES!!! - Bathroom-scale might be a few kilos out per hundred kg, so say +/- ~2%. Steel-ruler should be good to less than 1 mm, but percentage error depends on frame stiffness (stiffer frame = more proportional error). Bubble-level accuracy is usually 0.5 mm/metre (it says so on the tool, and easily verified). Half-track (with a "narrow-thing-under-the-tyre") should be good to +/- 1 cm, so +/- ~2%. And a few more potential error sources, which all students should look for...

Bottom line, the above half-hour (?) test can give you REAL, MEASURED torsional-stiffness that is LESS THAN 10% wrong. So it is 90+% right! And that is a lot more accurate than the imaginary numbers that come out of your FEA boxes. :)
~o0o~

INCREASING FRAME TORSIONAL STIFFNESS.
==============================

Bit briefer with this one.

Buy one 8' x 4' (2.4 x 1.2m) sheet of galvanised steel ~0.5 mm thick (or similar amount of 0.6 mm thk "zincalume", as used for roof-flashing, gutters, etc.). Cut to same shape as floor-rails (+~1 cm wider than centrelines), using manual tin-snips or electric-nibbler. Massage sheet so it fits snugly against all frame-tubes. Drill many ~3 mm holes in sheet where it contacts tubes, with ~2 cm to 10 cm gaps between holes (more holes in higher load areas).

With sheet clamped to frame-tubes, use MIG with spot-welding-timer to fill holes with a ~half-second blob of metal. Practice first on some scraps, but basically put MIG wire into hole, turn face away (you don't even need a helmet!), pull trigger, then repeat.

Maybe do same with selected side-panels of the frame (eg. the bottom-most 0.3 m, such as SIS and footbox area).

This sheet weighs ~4 kg per square metre, and you only need about one square metre (only ~half-sq.m if only doing the floor). So additional mass = ~ 4 kg, which gives you a very strong floor, and less side-bodywork.

Measure the frame's before-and-after torsional stiffness, and see it skyrocket! :)

Z

[mech5496]

As a side note on Z's post, that's how the Locost crew does things for 30+ years now.

[Charles Kaneb]

Z, one thing to add to the force drawing in your head on this measurement - springs representing the wheel and tire!

I remember being appalled at a TTC result showing a spring rate of well below 1000 lbs/in and then seeing a very similar result on a quick test I performed using a sheet of plywood and a ruler.

[Z]

... springs representing the wheel and tire!

Charles,

I am not sure what you are getting at, but I accept that the 3rd paragraph of my point 6 above was not entirely clear. So here is a clarification (and said paragraph now edited).

When doing the above Torsional-Stiffness testing, the car's spring and tyre-rates matter NOUGHT!

The "three wheel support" of the car puts the frame into an asymetrically loaded state, which amounts to a torsional Couple at one end of the car, equilibriated by an equal and opposite Couple at the other end. (And also vertical downward gravitational forces, equilibriated by two vertical upward forces, at F & R of the car, acting through the car's centreline. Err..., best seen by drawing an FBD.)

This applied loading, and its variation as more vertical load is applied to the driver's seat, causes the whole car, on average, to Heave, Pitch, and Roll with respect to ground, because of the car's squashy springs and tyres. But, again, this matters nought, because...

At each given "Torsional-Load" (= bathroom-scale-reading x T/2), you measure the "slope-of-front-bulkhead-top-tube" (= F-degrees) AND "slope-of-rear-bulkhead-top-tube" (= R-degrees), with both these slopes being wrt the local gravitational field (ie. wrt a level bubble-level).

Then ...[drum roll]... "Frame-Twist" is simply = F-degrees MINUS R-degrees.

If very stiff frame, then F-degrees always very close to R-degrees, although both can be large due to body-Roll from soft springs/tyres.

If very floppy frame, then F-degrees very different to R-degrees, because much frame-twist, even with rigid springs/tyres.

Z

[Adman]

Torsional-Stiffness (axle to axle of a car) is measured in kN.metres/deg (or radian).


That's really awkward... first year mistake. How embarrassing :(

We'll have a go at those methods and see what we come up with. I imagine it will be about 70-80% of the FEA projected figure.



When applying your constraints, where is the roll force being reacted? It's not at your wishbone mounting points. It's at your shock mounting points. Wishbone mounting points react lateral and longitudinal loads.

Thanks for the comments Menisk, that's very helpful. It's true we have not looked enough into what our chassis demands is terms of TS. I should be paying more attention to the calculations done by our suspension lead, as I cannot remember what he has assumed for the chassis 'spring' stiffness (if at all!).

While I was applying constraints to the front suspension loading points on the front, having just checked the FEA for the rear constraints I have indeed placed them too far back (on last year's chassis, anyway). It steps the design up from 900Nm to about 1250Nm, still less than half of our projected value of 2.7kNm for this year.

[MCoach]

I opted to make all of the Kettering 2014 and 2015 panels structural except for the access panel to pedals after we had rocked the standard "cover half the front end" style nosecone for decades. We did some chassis torsion tests and found that the riveted side panels alone carried 35-40% of the total stiffness for the 2014 car. This lead us to change some geometry, shorten up the chassis, increase side panel coverage, and still increase stiffness, despite the fact that we removed almost all non-regulation tubing and triangulation from the frame structure, saving about 15lbs. For comparison, our panels were 0.032" aluminum and weighed 8.5lbs total (3.8kg) total, so not far of from Z's weight estimate, however I would say that one square meter might be an underestimate... You could lose some weight and increase stiffness further by moving to bonded carbon or a bolted in carbon panel such as been seen from Monash or Zips (Akron) respectively, I believe. Considering the target market and cost mindedness of our team, we opted not to go that route.


Z, our baja team moved from using aluminum to steel panels for their belly pan. They very quickly returned back to aluminum because of the low impact resistance of the nearly atom thick steel they were using. I want to say it was 12 - 16 thou (.3 - .4mm). Food for thought.

[Ahmad Rezq]

Mcoach,
If I'am not wrong, I came across some of your 2015 pictures on your Facebook page, there were some suspension pickup points in the middle of frame tubes.
I'am thinking of adding side panels (1.2mm steel) to footbox area and remove all the additional tubes (keeping Only the 3 tubes required for the bulk head support).
Comparing the weight it added only 2 kgs compared to a well triangulated side. I will do some simulations maybe next week.

[Z]

This lead us to ..., shorten up the chassis, increase side panel coverage, and still increase stiffness, despite the fact that we removed almost all non-regulation tubing and triangulation from the frame structure...

MCoach,

That is the direction I would go if doing a "Rules-spec" spaceframe. Namely, structurally a monocoque, but you get it through scrutineering as a simple spaceframe.

Your Baja Team's use of 0.3 - 0.4 mm sheet-steel is at the lower limit of what is feasible for a floor getting that sort of abuse. The very common 0.6 mm sheet-steel used in Oz for roofing (ie. gutters, flashing, and available pre-painted as "Colourbond") is plenty strong enough. By comparison, I reckon there is very little sheet-steel on any Honda car out there that is much thicker than 0.4 mm.
~o0o~

As a BTW, the history of the "unit structure" car chassis-body is interesting.

IIRC, Mr Lancia started it in the early 1920s with his Lambda. But the idea didn't catch on, because while all-steel bodies were known to work well, it was clearly impossible for that flimsy stuff to also take the chassis loads. The Budd Manufacturing Company in USA kept promoted the idea heavily throughout the 1920s, but still no takers. Incidentally, the Chief Engineer at Budd, and the chief proponent of unit-construction, was a Mr Joseph Ledwinka, cousin of Hans Ledwinka, the CE at trendsetting Tatra.

But then a Mr Andre Citroen from France visited Budd, with his young genius CE Andre Lefebvre in tow. And ... in mid-1930s Citroen released its revolutionary "Traction Avant" (= "Front Drive"), the first mass-produced front-wheel-drive car with an all-unit-construction chassis-body. This car quickly developed a reputation for such good handling and performance that it became the favourite getaway car for bank-robbers, a bit like Jaguars in Britain a few decades later. Ahhh, praise doesn't come any higher! :)

So, it is possibly only because of the good sense of all those bank-robbers in the 1930s that the rest of the auto-industry eventually accepted that building a car's chassis out of paper-thin sheet-steel might just work. And now almost all of them do it! (Only light trucks still have separate chassis, and $upercar$ use CF.)
~o0o~

Ahmad,

1.2 mm is too thick. 1.0 mm is more than enough. And it is enough to do just the floor and two sides up to top-SIS-tube (ie. 0.3 metre high).

I will add more on your build thread.

Z

[MCoach]

Mcoach,
If I'am not wrong, I came across some of your 2015 pictures on your Facebook page, there were some suspension pickup points in the middle of frame tubes.
I'am thinking of adding side panels (1.2mm steel) to footbox area and remove all the additional tubes (keeping Only the 3 tubes required for the bulk head support).
Comparing the weight it added only 2 kgs compared to a well triangulated side. I will do some simulations maybe next week.

I believe there were only two pick up points that were not at a node on the chassis. I found the picture I think you were referring to.
For the front,he front lower points technically aren't at a node, but they're about an inch away from the 2 lateral tubes that support the steering rack that are not present in any of the publicly posted pictures.
I was never really happy with how they were hanging below the chassis, but there's more to the story there. So, in reality I guess nothing really is in the middle of a tube, but slightly offset, taking the bending load penalty (huge still!) to simplify a lot of stuff and get it out the door.
https://www.facebook.com/KetteringFSAE/photos/pb.209138079117577.-2207520000.1442671181./965187713512606/?type=3&theater


The rear control arm links pick up on the roll hoop like to Auckland's car. It's still partially braced on the front by the seat belt mounts, so there's that.
Reasoning was that the main hoop (reckon we have the biggest main hoop tube in the league) could take it easily.

For comparison at the rear:
https://www.facebook.com/KetteringFSAE/photos/pb.209138079117577.-2207520000.1442671802./999514596746584/?type=3&theater

https://www.facebook.com/UoAfsae/photos/pb.155875141099200.-2207520000.1442671565./604278989592144/?type=3&theater

[Z]

This quote taken from the Competitions section, but I think it is more appropriate here.


Originally posted by Adam:
I would like to believe that we have over engineered our car on the premise of reliability and ensuring we finish events (we will probably weigh close to 270kg without driver) but I'm sure we will break things along the way. 2016 will definitely be a year of refinement and tuning however. The last 2 years have been very scrappy (i.e. car running only weeks before comp).

Adam,

The above quote, written with your 2015 car barely running two weeks before comp, is IMO proof positive that UTAS have spent this whole year going in the WRONG DIRECTION.

Saying that "..2016 will definitely be a year of refinement and tuning" suggests that 2016 will be another year going even further in that wrong direction.

What will you "refine and tune"? A 270 kg heavyweight, with 3 unnecessary cylinders, an unnecessary (and probably understrength) machined-rear-bulkhead, umpteen unnecessary wishbones-with-push/pullrods&rockers+++, plus who knows what else you ADDED to the 2014 car?

And how will you do that "refining"? By ADDING even more machined-aluminium-bulkheads, maybe ADDING some carbon-fibre/3-D-printed-titanium whatsits, ADDING more etc., etc...?

The 2014 car had too much junk on it (~220 kg), but with some development and early build finish it might have been a realistic top-five contender this year.

If, at the beginning of this year, you took the 2014 car and greatly SIMPLIFICATED it (thus "adding more lightness"!), and if you also rearranged a few of the essential bits for better overall packaging (ie. more R%!), and maybe only added (*) a simple but effective aero-undertray, and got it all built early for lots of testing and tuning, then IMO it definitely would be a top-five contender this year. Depending on how well the testing and tuning went, and how good your aero-guy was, maybe even outright first-place contender.

So the tough question is, what will you "refine and tune" next year?

Which direction to go?

Add, or subtract?

Z

(* Edit: Good simulation shows that good aero is definitely worth "adding", especially to a lightweight car. It is a first-order performance improver. The other "added" stuff mentioned above is close to worthless. Or worse...)

[Claude Rouelle]

The heavier a car is the heavier it will have to be. The lighter a car is the lighter the car can be.

How in the hell is it possible today to design and manufacture a 270 Kg (without driver) Formula Student car? Why would you do this to yourself ?

[Menisk]

The heavier a car is the heavier it will have to be. The lighter a car is the lighter the car can be.

How in the hell is it possible today to design and manufacture a 270 Kg (without driver) Formula Student car? Why would you do this to yourself ?

This post isn't productive in the slightest. Go be a bully somewhere else Claude. These guys are engineering STUDENTS. For all you know it's a team of second year students that are only just starting to learn some basic engineering analysis. They're trying their best. Be constructive or don't say anything at all.

[BillCobb]

Anybody have statistics (median) for weights of cars by engine config? Might be a good place to start. Why would the weight stats NOT include the driver?
Just curious.

[apalrd]

If you look at students.sae.org for results (http://students.sae.org/cds/formulaseries/results/) and download the file, the last tab has 'info' which includes team country, engine cylinders, displacement, and vehicle weight without driver. I haven't looked for data from other competitions.


Considering only 2015 Lincoln, the mean weight is 178.32kg for 4cyl cars, 190kg for 2cyl cars (only one has data!), and 165.4kg for 1cyl cars.

[BillCobb]

The old hidden off the deep end INFO tab trick.

[ritwikdas18]

Well Claude is right. Even Indian teams like us manage to build a car at ~210kg in our second year with zero funding from university, lack of machining facilities and other problems.270kg is on the higher side even to the standard of Indian teams so for teams with much greater facilities and resources at their disposal 270kg is not that great.

[Claude Rouelle]

You can't make omelets without braking eggs. Similarly there are no sweet ways to say to a team that the best they try is still not good enough.

270 kg is simply not acceptable even less for a second year car.

If students create a car at 270 Kg that probably means that minimum weight was not part of their target list and/or that they probably did not know from simple and basic simulation how much weight influences the car performance.

Formula Student or FSAE is an engineering and project management challenging educative program. For ecological, financial and even political reasons (energy consumption, pollution while using and manufacturing vehicle, recyclable-ability etc...) the whole mechanical engineering industry (train, car, aircraft, aerospace) is looking for new techniques to make things cheaper, lighter and stiffer. Designing and manufacturing a car at 270 kg shows that the team did not use "the form follows function" option and the students did miss a huge chance to get best prepared for their future job.

It is possible to design and manufacture a stiff, light, reliable, safe, easily passing technical inspection, cheap to manufacture, easy to maintain, 4 cylinders, 13 in, tubular chassis, with wings under 175 Kg, definitely under 200 Kg. I have seen it being done by teams the very first year with low budget, no real racing or automotive engineering education or culture in their country, no amazing education level and poor support form their university staff. I still could find some words of encouragements at 220 Kg for a first year car if I see some good design or innovation or good manufacturing of specific car parts but do not expect praises for me for a 270 KG second year car.

Without any additional explanation I would probably believe that a team that manufactures a 270 Kg car either did not gather the basic information that explains what makes a car competitive (and that information is easily available) or simply do not care.

Am I harsh? You bet! But wait not a faculty adviser or a design judge but for a real boss, a company with customers, competitors, deadlines and budget constraints.

[Claude Rouelle]

You can't make omelets without braking eggs. Similarly there are no sweet ways to say to a team that the best they try is still not good enough.

270 kg is simply not acceptable even less for a second year car.

If students create a car at 270 Kg that probably means that minimum weight was not part of their target list and/or that they probably did not know from simple and basic simulation how much weight influences the car performance.

Formula Student or FSAE is an engineering and project management challenging educative program. For ecological, financial and even political reasons (energy consumption, pollution while using and manufacturing vehicle, recyclable-ability etc...) the whole mechanical engineering industry (train, car, aircraft, aerospace) is looking for new techniques to make things cheaper, lighter and stiffer. Designing and manufacturing a car at 270 kg shows that the team did not use "the form follows function" option and the students did miss a huge chance to get best prepared for their future job.

It is possible to design and manufacture a stiff, light, reliable, safe, easily passing technical inspection, cheap to manufacture, easy to maintain, 4 cylinders, 13 in, tubular chassis, with wings under 175 Kg, definitely under 200 Kg. I have seen it being done by teams the very first year with low budget, no real racing or automotive engineering education or culture in their country, no amazing education level and poor support form their university staff. I still could find some words of encouragements at 220 Kg for a first year car if I see some good design or innovation or good manufacturing of specific car parts but do not expect praises for me for a 270 KG second year car.

Without any additional explanation I would probably believe that a team that manufactures a 270 Kg car either did not gather the basic information that explains what makes a car competitive (and that information is easily available) or simply do not care.

Am I harsh? You bet! But wait not a faculty adviser or a design judge but for a real boss, a company with customers, competitors, deadlines and budget constraints.

[Ahmad Rezq]

It is possible to design and manufacture a stiff, light, reliable, safe, easily passing technical inspection, cheap to manufacture, easy to maintain, 4 cylinders, 13 in, tubular chassis, with wings under 175 Kg, definitely under 200 Kg. I have seen it being done by teams the very first year with low budget, no real racing or automotive engineering education or culture in their country, no amazing education level and poor support form their university staff.

Claude,
It will be very helpful if you can mention some of those teams.

[P^squared]

Well Claude is right. Even Indian teams like us manage to build a car at ~210kg in our second year with zero funding from university, lack of machining facilities and other problems.270kg is on the higher side even to the standard of Indian teams so for teams with much greater facilities and resources at their disposal 270kg is not that great.

What we're missing out here is the ~35 kg heavier powertrain, ~15 kg ?? in unsprung aero, and other extra sub-systems.
( Not implying all of those weight gains are defendable ! )

My point is that our 210 kg singles are in the same league as UTAS' car ! :D

The need for machining facilities to make a lighter car is definitely debatable !

[Claude Rouelle]

Amadh,

You can find this information on the web. That being said what, how useful is it for you to know which team did create a car of which weight?

I ask this because I see too many students rushing on the internet to see "what the other do", instead of designing their own stuff THEN and only THEN compare their design solutions with the ones of their competitors. By looking at the design of other cars they get themselves out of the process of defining the functions before defining the forms and their restrain their imagination.

If you allow me a larger perspective the goal of this competition is not to win; it is to do the best you can. Of course you want to win and I admire and i congratulate you for that. But do not reverse the priority.

I believe in inside-out solution (what do we think a car need to be competitive? - Can we as a team do that? - What are our resources? - How can we expend those and efficiently use them- Let's put our imagination at work) more than outside-in solutions (what did the winner do? - Let's copy that!)

I know several students who do really well in Motorsports and who created FS cars that, 5 or 6 years ago, were not very impressive on track. I have seen winners that did not do the best they could and I have see students who did not created impressive cars but did the best they could. Of course you want both but which ones do you think prepare themselves the best for an engineering career?

[Claude Rouelle]

Amadh,

You can find this information on the web. That being said what, how useful is it for you to know which team did create a car of which weight?

I ask this because I see too many students rushing on the internet to see "what the other do", instead of designing their own stuff THEN and only THEN compare their design solutions with the ones of their competitors. By looking at the design of other cars they get themselves out of the process of defining the functions before defining the forms and their restrain their imagination.

If you allow me a larger perspective the goal of this competition is not to win; it is to do the best you can. Of course you want to win and I admire and i congratulate you for that. But do not reverse the priority.

I believe in inside-out solution (what do we think a car need to be competitive? - Can we as a team do that? - What are our resources? - How can we expend those and efficiently use them- Let's put our imagination at work) more than outside-in solutions (what did the winner do? - Let's copy that!)

I know several students who do really well in Motorsports and who created FS cars that, 5 or 6 years ago, were not very impressive on track. I have seen winners that did not do the best they could and I have see students who did not created impressive cars but did the best they could. Of course you want both but which ones do you think prepare themselves the best for an engineering career?

[Claude Rouelle]

P^Squared (what is your name / team by the way / country?)

210 Kg mono-cylinder is not amazing either. You car is effectively in the same league as the 270 Kg 4 cylinders except that you should have a lesser energy consumption

Most of the students keep thinking about making a lighter car. Lighter than what? What is the reference? Your team last year car? Last year winner?

The goal is not to reduce last year car by x Kg. The goal is not to "add lightness" but to start from a blank sheet of paper from 0 Kg and to increase the weight to the minimum.

That is where the "Form Follow Function" principle - not the other way around - comes from. I am not a very religious person but if we look in nature at God creations (a snail shell, a wale skeleton) and understand Darwin's perspective, we understand that it is that principle that will help you to make a car both light and stiff and easy to manufacture.

[Z]

If you look at students.sae.org for results (http://students.sae.org/cds/formulaseries/results/) ...

Considering only 2015 Lincoln, the mean weight is 178.32kg for 4cyl cars, 190kg for 2cyl cars (only one has data!), and 165.4kg for 1cyl cars.
Andrew,

You might want to check and edit that above post, just in case any newbies take it as gospel. (Edit: Maybe your averaging included all the 4-cylinder cars with "0 kg" weights?)

My quick look through Lincoln-2015's numbers (which are noted as "not validated") gives from lightest car up,

143 kg - 250 cc Univ Cal-Berkeley,
155 kg - 450 cc Kettering,

... then for 600 cc/4-cylinder cars only,
175 kg - Kennesaw State U (?*), then,
190->210 kg - 10 x cars,
211->230 kg - 20 x cars,
231->270 kg - 14 x cars,
271+ kg - 3 cars.

The three lightest cars named above all scored well, ... until Endurance, where negligible score!

But I reckon Enduro failures are more often caused by excessive complexity, rather than excessive "lightness", in itself.

Z

(*PS. From a quick look at Kennesaw's facespace, they have a conventional looking 600 cc/10"/spaceframe car, quite neat, and with Direct-Acting-SDs (albeit too high rear RC). But I find it hard to believe they are ~20 kg lighter than the other lightweight 600/10"/SF cars?)

[Claude Rouelle]

I think I told that story in this forum before but I guess it is worth to tell it again
2011 FSG – E. I judged these 2 electrical car teams. Both cars had tubular chassis, 4WD, no wings. One was 80 KW and the other 82 KW. So very similar concepts.

One car was 145 kg and the other 234 kg. When I asked the “heavy” car team what did justify a 50 % weight difference to their competitor for the same concept, they told me that their car was heavier because it had more batteries. I asked why they had more batteries and they told me that without it they could not finish endurance. I asked again why they needed that more batteries and they told me that was because… their car was heavier!

Some students do mix causes and consequences.

The light car had different top and bottom wishbones and toe link rod ends. The heavier car had 8 mm rod end everywhere.

Again: The heavier the car is the heavier the car will need to be. The lighter the car is the lighter it can be.

When I discussed with the 234 kg team their engineering approach, they told me defensively that their car was 30 kg lighter than this previous year: I was not convinced by that argument. If there is a moment where you do not work in “Delta” that is at the moment of the car concept phase.

Again: The goal is not to reduce last year car by x Kg. The goal is to start from a blank sheet of paper from 0 Kg and to increase the weight to the minimum

For the same performance of let’s say 1.5 G lateral on a skid pad the 4 tires of a 200 kg (with driver) car will generate a total of 3000N of lateral forces.
For the same 1.5 G a 300 kg car (also with driver) the 4 tires will generate a total of 4500N of lateral forces. F = ma and action = reaction.

You can be the king of the FEA and the master of design but the heavier car will inevitably see more possible compliances (more red than blue) than the light car. This compliance makes the driver uncomfortable. It most often decreases response and stability. Just imagine the number of springs, dampers and hysteresis you have between a steering wheel and the 2 front contact patches and how the driver feels about that.

So you want a light and stiff car that is responsive or a heavy, lazy car that twists and bends like a piece of rubber? What if the reason why a car is bending is its weight?

When you have a headache you take an aspirin. It works. You worked on the consequence. That is OK if you have 1 or 2 headaches a year. But if you have a headache every day you need to find the cause and work on it.

If you have rear toe compliance the “aspirin’ solution is to use stronger and heavier wishbones and rod ends. The weight will increase and the crescendo to the disaster will go on.

The smart solution is for example to reduce the causes of the compliance by increasing the toe base or decreasing the Mz with the right choice of tire KPI and mechanical trail.

It is possible to design and manufacture a light and stiff car. Actually it is easier to make a light car stiff.

To do that you just need to think Function then Form not the other way around. But that requires a bit of imagination and abstractive thinking ideally in group before you rush to the CAD software. These skills are within each of us and won't be found on the internet.

[Jonny Rochester]

OK I think it's time I step back in here to represent UTAS. Since our car is (a bit broken) on the dyno, I have a chance to sit down and explain the situation. Just let me get a coffee.

[Jonny Rochester]

Spec sheet reminder:
CBR600RR, ~2009
13" x 7" wheels/ tires
4130 tube chassis
carbon undertray and wings
fibreglass and cork body

Firstly, UTAS are now rolling, and we have driven the car in 1st gear up the laneway and tried some basic noise and brake tests. This is all in the last few days. From driver feedback and what I saw, this car has all the potential of any 4cyl FSAE car. (Yes, we are out of time for extensive testing and development, you don't have to tell me!!!)

We only bolted wheels to it and put it on the ground about a week ago. But having last years experience we are able to correct small oversights quickly.

When we lifted the car with a chain off the roll hoop, we noticed it was almost "perfectly" balanced on the chain. That is, when hung from the main roll hoop, the car sits horizontal. This is with no impact atenuator and no aero. I will leave you to argue if that is "good" but it is about what I wanted and expected. The driver will bring this point forward, but we should have a bit of R% bias rather than 50:50.

With the car on the ground on 4 scales, the numbers where all over the place as I expect the corner weights to be wildly off, as well as having the scales on a uneven floor and not adjusted and a stiff chassis also contributing to this. Anyhow, in the time we had we estimated it was 220kg as a drivable car, but without aero. There has not been any focus on weight in our team in the last 6 months. The weight is just what it is (don't quote me on any numbers yet), we know it's more important to complete it.

With COW done (car on wheels), we had to work fairly quick to get the wiring harness done. I had a late night. It's mostly working but I'm going to redone some of it when I get a chance. Some power wires where getting warmish.

I got stuff done just enough to we can start the motor and sort of drive it a bit. No time to attempt tuning the motor, it ran so that was good enough. Then we put the car in our brand new trailer and took it to the dyno tuner. Tasmanian people will know of PRO Automotive. We got a phone call back to say we have at least 70hp, got to 150km/h in 4th, but our plastic intake plenum cracked a bit. So that's where we are at at this stage.

[Jonny Rochester]

Some basic failures in the last 2 days:
pin hole leak from in-house fabricated radiators, easily rewelded
3D print throttle pedal broke, was a prototype anyway, CNC aluminium one on the way.
plastic plenum cracked, we are gluing it now and will wrap with carbon
other air leaks in plenum, now using sealant.
movement in our steering uni joint, trying to order a stronger one.
wings and body yet to be mounted.
And of course noise is too noisy! ~120dbc

860861862863


I made the exhaust. It's not in CAD. The lengths are perfect, but only I am allowed to measure it. It is now ceramic coated in black as seen.

[Menisk]

It is possible to design and manufacture a stiff, light, reliable, safe, easily passing technical inspection, cheap to manufacture, easy to maintain, 4 cylinders, 13 in, tubular chassis, with wings under 175 Kg, definitely under 200 Kg. I have seen it being done by teams the very first year with low budget, no real racing or automotive engineering education or culture in their country, no amazing education level and poor support form their university staff.

This is a load of bullshit. If you're going to make a comment like this back it up with an example. Find me a team that built a 13" wheeled, spaceframe chassised 4 cylinder car with wings for under 175kg and finished endurance. In Australia ECU is building some of the lightest 4 cylinder cars in the world, they're in the mid 180's. It's a carbon monocoque and rolling on 10" LC0's (the lightest tyre that's commonly run on these cars) with a custom engine. Having had a good look over their first iteration myself, I can tell you it's an impressively engineered machine.

I have a huge problem with the way you deal with people on this forum. There are other people on this forum that can be quite controversial, for example Z ruffles feathers all over the place, but every time he ruffles feathers he offers something productive and he will continually reply and try to work with people to help them achieve a better understanding. I'm not saying I completely agree with the way he interacts, but I can respect him for the fact that he doesn't care who you are and where you're from, he'll try to teach you if you show that you want to learn. I won't deny that you do attempt to post productive things and promote learning, but there's so many posts on this forum from you that are just plain bullying. You feel the need to confront the new guy that's asking a legitimate question about where he's from and who he is. It's not important! If he has something interesting to discuss and people want to discuss it with him, let them. Asking people for their credentials and trying to gain contexts is akin to prejudice. Should we start putting our race, gender and age on our posts too?


Am I harsh? You bet! But wait not a faculty adviser or a design judge but for a real boss, a company with customers, competitors, deadlines and budget constraints.

Telling people that their work is shit is a reality of life and it needs to happen if their work is going to improve. I won't argue that. However when you offer nothing productive along with your criticism you achieve nothing productive. You just put someone down for the sake of it. Why would you do this? What do you hope to achieve? Does stomping all over students that are trying their best make you feel good about yourself? You forget very easily that we're STUDENTS. You have no idea of the level of experience of the UTAS team. You have no idea how many members continued over from last year or how well the previous team members attempted to transfer knowledge. I don't know the entire situation either, but there will be plenty of teams around the world being started by a handful of second year students that haven't done half the courses they would need to in order to design a good car. So they work with the knowledge they have, add plenty of safety and do their best.

You need to stop treating engineering STUDENTS like graduates. Lots of us do this alongside a full uni load and a life where we have to balance the commitments of family, relationships and work. If we don't have time to learn every detail about FEA, Aerodynamics and Tyres then you'll forgive us for trying to design cars with the best knowledge and limited time we have.

[Claude Rouelle]

Menisk

In life in general and in racing in particular there are two categories of people: one that makes winning and the one that makes excuses. We have to decide which category we want to belong to.

Claude

Claude

[Rex]

This is a load of bullshit. If you're going to make a comment like this back it up with an example. Find me a team that built a 13" wheeled, spaceframe chassised 4 cylinder car with wings for under 175kg and finished endurance.
Not just a car meeting those criteria. One that was also built in a shed, with a hatchet, by underprivileged orphans, in the snow, uphill both ways. Remember, the point here is NOT that such a car could theoretically exist and you should strive for it. The point here is that anyone with half a brain and no resources can build (and indeed already has built) such a car, so why the hell can't you?


I have a huge problem with the way you deal with people on this forum... I won't deny that you do attempt to post productive things and promote learning, but there's so many posts on this forum from you that are just plain bullying... Telling people that their work is shit is a reality of life and it needs to happen if their work is going to improve.... However when you offer nothing productive along with your criticism you achieve nothing productive... You need to stop treating engineering STUDENTS like graduates.
Agreed, with one small adjustment. Even for graduates in the real-world, I consider many of Claude's interactions unacceptable. I run a fairly successful business, and I can tell you that if any of the managers in my organization treated people the way Claude treats people on the forums, they would be fired. Contrary to what Claude's stated philosophy would lead you to believe, not everyone in the real world is a jerk.

[Menisk]

Menisk

In life in general and in racing in particular there are two categories of people: one that makes winning and the one that makes excuses. We have to decide which category we want to belong to.

Claude

Claude

I'd like to belong to the category where people are pleasant and considerate of each other.

[Adman]

Oh boy... what have I started here. Maybe I should just say the 270kg was a typo! ;) Relax, the car is not 270kg, but until recently we had not had an accurate weigh-in. I'm not going to add fuel to the fire here as most points have been made, but if I may, lets set some facts straight so people know a bit more of our story
- we are not one of the "big" uni's in Australia, and so we tend to attract a lot of local Tasmanian's who do engineering "because". Think of it like the African brain drain... most of our brightest students transfer over the mainland after 1st year. We get many of these joining our team, and we had 42 people back in January. We are now down to 18. 16 heading to comp. It makes it very difficult to judge where to spend resources, time, effort and place students in areas. Suspension is now a one man team, which leads me to my next point...
- all students complete honours, or a final year project, of which this year we HAD 4. The driveline guy decided to leave in June with unfinished designs. He was slightly ambitious, but had done plenty of analysis so we didn't have to do too much to finalise his designs and get them made. Our aero leader wanted to design a base-line aero kit that could be expanded upon in years to come. This was her passion, she worked INCREDIBLY hard, but has been let down by members of her team leaving (and perhaps lack of communication). Suspension was again a 4 man team in January, it's now a single guy who's final year project was to develop and analyse the same suspension system we had last year, and his decision after many many matlab & simulink calculations was to develop ARBs. It is VERY difficult to tell these guys that their projects cannot go ahead... and this is why we have some additional components on the car this year, vs simplifying old ones.

Now,

Z : I like your points, most are valid. I wouldn't say we have gone in the wrong direction, but we have certainly overshot what we could have done. Again, as I've said, this was due to an massive influx of students in January (way more than we had in 2014) and so we thought we had to capability to complete what we set out to achieve.

Claude : Some just criticism, some unhelpful comments. As Menisk mentioned, you have no idea of our budget, facilities, experience, knowledge transfer. I think we have a major team management problem, and is something I will be asking around the teams at competition to fix for next year. Our faculty advisor loves the bureaucracy of several team leaders, each with team members that report upwards in the chain etc. But also, the only comment he could make when i designed our 2.8kg engine bulkhead (with extensive ANSYS FEA) was that "[he] would be happier if it weighed closer to 5kg". So I guess we're lucky the car only weighs 270, as it might've weighed 272.2 if I'd followed his advice!

To ritwikdas18 & P^squared; thanks for you criticism, not very constructive or supportive, but anyway. We do not have many of the benefits of mainland unis, mainly a lack of major industry support (though we do have a nice workshop). Also very poor university funding. But we manage without.

__________________________________________________ ____________________

A separate point;

Having been lurking on these forums since the end of 2013, I was surprised to see a lack of 'build threads'. I thought it would be a common thing (as you would see on many car enthusiast forums), but oddly not. Looking at many of the comments in this thread, it's pretty obvious why.

Utas has put itself out there for critcism, as have Jon and I. It's a good thing, else we would not bother! But there is a serious problem with the community here as Menisk has mentioned, intrigued students are treated as inferiors and their ideas often shunned.

__________________________________________________ ____________________

I'm gonna back out for a bit, and grab some more popcorn.

[Mitchell]

If you can design/build a car in a year, complete competition events, pass all your uni courses, learn some useful knowledge and become a better engineer you have succeeded at fsae. Doesn't matter what the car weighs, or what trophies are associated with it.


It is possible to design and manufacture a stiff, light, reliable, safe, easily passing technical inspection, cheap to manufacture, easy to maintain, 4 cylinders, 13 in, tubular chassis, with wings under 175 Kg, definitely under 200 Kg. I have seen it being done by teams the very first year with low budget, no real racing or automotive engineering education or culture in their country, no amazing education level and poor support form their university staff.

Claude, I understand that this is a complete lie and intended as a motivational comment. Maybe next time though make it a tad more realistic.

[Kevin Hayward]

Adman,

Your team size sounds very familiar to ECU. I think we have 17 going to comp this year, one of which will be a final year, not an unusual outcome for a small team.

Solid mentoring and more flexible management structures can allow this to work pretty well. A big focus on having 1st and 2nd years in decent team roles really helps here.

Simple fact is that you don't need a team of 60+ to do well in FSAE. Some of the judges have a funny idea of what is required and should be expected of all teams, but the beauty of being a small team is that you can only afford to focus on the essentials.

While our guys might be envious of teams with beautiful videos from marketing teams larger than our design team, it doesn't make too much of a difference with the quality and performance of the vehicle.

Hopefully you guys will have good continuity next year.

Kev

[Adman]

So Utas achieved 15th out of 30 this year. It's a little lower than we were hoping for, but I think it's an expected result.

We suffered for sure in not getting enough testing done before comp; the car was poorly set up for autocross (and undoubtedly for acceleration and skidpad, although the problems weren't as obvious in these). Myself and the suspension leader were drivers for autox and neither of us were able to make the hairpins and sharpest corners. A bit disappointing as we feel the car has high potential, we just couldn't bring it out.

Some sunday morning adjustments transformed the car and made endurance a far more enjoyable event. For some reason we got 0 for economy despite completing the first endurance, with our first pair of drivers setting times around 2:00. Our faster drivers were out in the second heat and set times of 1:48 - 1:50, and perhaps we could have been faster with better setup. Unfortunately I managed to snap the throttle cable a few laps after the driver swap... ending our chances of a redemption from the autox event.

The 'would've, could've, should've" in me believes that with an improved setup we could've achieved our goal of a top 10 finish. Our static events improved considerably this year (5th in presentation, 9th in design, a sore 17th in cost).

Despite not running our aero package, Riccardo (GTS, apologies for a potential mispell) seemed impressed with our aero leader's solo attempt at a bodywork, undertray and wings combination. Quote "should be doing a PHD in aero" and semi-quote "10 years of work from Monash undermined by a single girl from Tasmania". We'll be working in the next week or two to get these fitted and tested in combination with our suspension tuning. Hopefully they live up to the expectations of GTS; you'll see them again next year with some small improvements I'm sure!

So... 2016!!

More bulkheads, FEA and CFD, plus all manner of CNC! Oops... what am I thinking.

Jokes aside, at this stage obviously things are only very vague, but essentially we aim to return with a similar concept car, although much refined. A bit of a diet, fully tuned suspension and aero, and reduced steering-effort are up there on the list.

Overall we really enjoyed the competition, was fantastic to see so many competitive teams (both old and new). We'll be back next year for sure! :D

[GTS]

Despite not running our aero package, Riccardo (GTS, apologies for a potential mispell) seemed impressed with our aero leader's solo attempt at a bodywork, undertray and wings combination. Quote "should be doing a PHD in aero" and semi-quote "10 years of work from Monash undermined by a single girl from Tasmania". We'll be working in the next week or two to get these fitted and tested in combination with our suspension tuning. Hopefully they live up to the expectations of GTS; you'll see them again next year with some small improvements I'm sure!

I'll qualify that.

PhD in aero - yes, your team member should at the very least apply for an APA (as should anyone with a strong interest and the grades to move forwards) - an APA scholarship is based around the student, and can be deferred to some degree. So there's flexibility.

Regards the semi-quote, the rear solution in particular was very intelligent and had a few other teams (Monash included) taking a peek. And that wasn't just my opinion, there were two judges and the other thought similarly. Some may say that Monash's solution generates more net load, I'd say you'd hope it does given the years Monash has been refining their solution - as a concept, though, the UTAS solution was very intelligent. A very polished concept can certainly be undone by an early attempt at a better one.

We also implored your faculty representative to run the parts.

For PC's sake "by a single girl from Tasmania" was actually "by one woman from Tasmania".

Not to undermine anyone at or associated with Monash - the program has produced some excellent grads and the three guys presenting this year look set to be no exception to what's been a very high standard - but what was achieved (and not run, incredulously) by UTAS was simply a more difficult thing to do.

Well done. This is, in part, what FSAE is about.
(So run the parts!)

[CWA]

Sounds like you've learnt a lot chap.. and that's what it's all about

Some things from me:
- Didn't last year's UTAS car also have a high steering effort issue?
- What does "run the parts" mean?

[MCoach]

Oh boy... what have I started here. Maybe I should just say the 270kg was a typo! ;) Relax, the car is not 270kg, but until recently we had not had an accurate weigh-in. I'm not going to add fuel to the fire here as most points have been made, but if I may, lets set some facts straight so people know a bit more of our story
- we are not one of the "big" uni's in Australia, and so we tend to attract a lot of local Tasmanian's who do engineering "because". Think of it like the African brain drain... most of our brightest students transfer over the mainland after 1st year. We get many of these joining our team, and we had 42 people back in January. We are now down to 18. 16 heading to comp. It makes it very difficult to judge where to spend resources, time, effort and place students in areas. Suspension is now a one man team, which leads me to my next point...
- all students complete honours, or a final year project, of which this year we HAD 4. The driveline guy decided to leave in June with unfinished designs. He was slightly ambitious, but had done plenty of analysis so we didn't have to do too much to finalise his designs and get them made. Our aero leader wanted to design a base-line aero kit that could be expanded upon in years to come. This was her passion, she worked INCREDIBLY hard, but has been let down by members of her team leaving (and perhaps lack of communication). Suspension was again a 4 man team in January, it's now a single guy who's final year project was to develop and analyse the same suspension system we had last year, and his decision after many many matlab & simulink calculations was to develop ARBs. It is VERY difficult to tell these guys that their projects cannot go ahead... and this is why we have some additional components on the car this year, vs simplifying old ones.

Now,

Z : I like your points, most are valid. I wouldn't say we have gone in the wrong direction, but we have certainly overshot what we could have done. Again, as I've said, this was due to an massive influx of students in January (way more than we had in 2014) and so we thought we had to capability to complete what we set out to achieve.

Claude : Some just criticism, some unhelpful comments. As Menisk mentioned, you have no idea of our budget, facilities, experience, knowledge transfer. I think we have a major team management problem, and is something I will be asking around the teams at competition to fix for next year. Our faculty advisor loves the bureaucracy of several team leaders, each with team members that report upwards in the chain etc. But also, the only comment he could make when i designed our 2.8kg engine bulkhead (with extensive ANSYS FEA) was that "[he] would be happier if it weighed closer to 5kg". So I guess we're lucky the car only weighs 270, as it might've weighed 272.2 if I'd followed his advice!

To ritwikdas18 & P^squared; thanks for you criticism, not very constructive or supportive, but anyway. We do not have many of the benefits of mainland unis, mainly a lack of major industry support (though we do have a nice workshop). Also very poor university funding. But we manage without.

__________________________________________________ ____________________

A separate point;

Having been lurking on these forums since the end of 2013, I was surprised to see a lack of 'build threads'. I thought it would be a common thing (as you would see on many car enthusiast forums), but oddly not. Looking at many of the comments in this thread, it's pretty obvious why.

Utas has put itself out there for critcism, as have Jon and I. It's a good thing, else we would not bother! But there is a serious problem with the community here as Menisk has mentioned, intrigued students are treated as inferiors and their ideas often shunned.

__________________________________________________ ____________________

I'm gonna back out for a bit, and grab some more popcorn.

If there is anything I've learned, it's do not to be afraid to be wrong. I've spent my entire life being wrong with very few select moments of being correct. The criticism that many have given, especially Z with his fundamental questions has constantly challenged us to go back and make sure we understand the basics, able to form individual thoughts, and provide goals for ourselves (however unrealistic some of them may be). Now, not everyone needs to be shouting, nor interacting in a condescending way, because we are all students. Students study the materials, teachers study the students.

If anyone told me where the FSAE community level of expertise and competitiveness would be in 5 years when I joined, I wouldn't believe them. Flat out, would not believe.

I guess the only thing else that I have to add to UTAS is that software and simulation are cheap compared to making physical objects. Carrying concepts from year to year that are considered "acceptable for now" can allow you to spend more time using and building analysis tools to decide what systems/parts are too heavy, what is too compliant, what doesn't fall under the team philosophy of building the best car you can muster. Simulation tools can allow you to make the decisions of what areas to tackle first and what you're limiting features are. Aero? Chassis weight? Suspension geometry? Component selection? As powerful as these tools can become, their accuracy is always limited by their assumptions and their usefulness is always limited by "time cost" of using the software.

Good luck and looking forward to what you may produce next year.

[Adman]

Sounds like you've learnt a lot chap.. and that's what it's all about

Some things from me:
- Didn't last year's UTAS car also have a high steering effort issue?
- What does "run the parts" mean?

You are correct regarding steering. For some reason or another, the same problem has occurred again. It is better mind you, but compared to driving other cars we are still behind. There may have been a miscommunication between myself on chassis and our suspension leader, as we are still not making full use of our steering rack travel which we purchased on a slower ratio to improve the effort.

"Run the parts" refers to the front and rear wings that we could not get mounted in time. GTS gave our faculty advisor a good flogging for it (which the team somewhat enjoyed) but in reality our aero leader hadn't fully considered how to mount it. It was always a "she'll be right" attitude. There was no cross bracing on front and rear wings (so it would just flop side to side) and little consideration was taking for the space taken by the bodywork and exhaust routing. Admittedly some things got in her way unexpectedly (like our CAD model for the tyres not being accurate).

We've learnt a lot, and aim to get the most knowledge retention we can over the summer to give our 2016 car the best shot it can.

GTS, thanks for your comments. It's awesome to have a judge that can back up their comments with feedback!

MCoach, again thanks for you comments. We have a solid idea about where we want to go this year, it's more about how we manage ourselves that we know has to change.

[GTS]

- What does "run the parts" mean?


It means stick them on a car, get some data and experience. The excuse given for not running the parts was... not good (it didn't come from the student involved). Simple application of Logic Sticks and Reality Helmets, ~20 odd students with a copy of SolidWorks (I'll assume all are legal installations!), some steel, welder and fasteners would have made for a directionally-correct solution. The other concerns mentioned are fixable... every project starts thus. Time, cost and quality, you can pick any two in any instant etc...



GTS, thanks for your comments. It's awesome to have a judge that can back up their comments with feedback!


Anytime. (Bonus points for correct spelling on your previous post).



In life in general and in racing in particular there are two categories of people: one that makes winning and the one that makes excuses. We have to decide which category we want to belong to.

A bit interesting to read further up.

Claude, I get this, having worked in racing, 'life in general' and the like. But I've also been faculty and, many moons ago, a Formula SAE student.

I sympathize with the notion that many efforts simply set poor targets. There are many reasons for this, some to lack of knowledge, most to lack of confidence. Some of the latter goes far deeper than just the students involved and well into faculty that, for better or worse, don't believe the collective efforts of their students and resource network are capable of winning (those present last year might remember a few of us having a rather memorable argument to this end). Whilst your science is in the right place, I'd suggest that where the issue can be confidence, tearing it apart in an open forum won't build it.

Nor is the ultimate goal of everyone's FSAE career winning. There's motorsport for that, and this isn't it.

[Claude Rouelle]

GTS,

More than confidence it is the lack of ability to understand and measure what is necessary to make a winning car and a winning team.

That you do not have the resource, the time,the budget, the education, the labs, the teacher, the software etc....to make a wining team and car, especially the the first year, is understandable: you need to start somewhere.

But not having a bit of imagination and search skills and take the time to create that list of winning criteria is an insult to your own intelligence and the people who have invested in you. (I mean "you" in general, not GTS in particular)

[GTS]

GTS,

More than confidence it is the lack of ability to understand and measure what is necessary to make a winning car and a winning team.

That you do not have the resource, the time,the budget, the education, the labs, the teacher, the software etc....to make a wining team and car, especially the the first year, is understandable: you need to start somewhere.

But not having a bit of imagination and search skills and take the time to create that list of winning criteria is an insult to your own intelligence and the people who have invested in you. (I mean "you" in general, not GTS in particular)

Claude;

I'll go the other way.

Whether in racing or in FSAE (which certainly isn't racing), self-belief is everything. Whilst agreeing winning isn't chance, citing 'lack of ability' reads poorly. These students were each smart enough to get into an engineering degree program, to which admission isn't free. They've all a basic command of Newton's 2nd and get the basics just fine. The number of students I've met that set goals behind the ultimate is staggering, but it's very little to do with not being able to identify winning criteria. In many cases the 'we set our targets mid-pack' statements really boil down to 'we're not confident we can do better', not 'we're not sure what better is'. Often the very attitude you cite is endorsed by the very faculty and stakeholders that students look to for guidance and endorsement. No student wants to come into FSAE, give up a good amount of free time, put their name to a prominent piece of work all in the name of targeting mediocrity at best - if that's an assertion it's a silly one.

If they were completely able to guide, research, self-endorse and actuate, they'd not be students, Claude, they'd be professionals. The role of this project is to accelerate this journey, not to critique that it is, in fact, a journey. The message needs to be you can do it, not you're an idiot for not having done it. It's a learning experience the first time, a mistake when made the second time - and there's plenty of career for the latter.

Teams - complete teams, not just individuals within - need to believe they can win and want it too, and that's not easy within the student FSAE environment. Nor is it the point of it.

Modern FSAE doesn't make it any easier, we give students industry-grade CAE tools and wonder how they can't deliver when in reality, these things are very complex and require a level of commitment and proficiency not possible in what remains a part-time student project. They don't actually know better that to drop complexity in execution increases reliance on their heads, which given free reign are particularly powerful things.

Our local competition has a few elements of it's own which remain a bit shitty to deal with. Some are political. On aero particularly it took a rule change to force free thinking. Nothing's perfect.

When individuals in the FSAE environment buck the trend enough to commit to their potential beyond any self-doubt and end up delivering something special, then it's a real achievement. To this end, UTAS FSAE achieved.

[Kevin Hayward]

Well said GTS

[Jonny Rochester]

Some things from me:
- Didn't last year's UTAS car also have a high steering effort issue?


Yes. I even measured the steering torque required when stationary, made note of the steering wheel diameter, steering rack ratio, and upright toe base. I/we listed all the possible contributions to steering effort, then tryed to change a few. I think the main one is the steering rack ratio combined with the toe-base length, but I also looked at the wheel alignment angles. After talking to another team I chose some very low numbers for KPI, caster and scrub, and our suspension guy went along with this after doing his own calcs. I would have increased toe-base also but it would seam now the suspension guy thought the wheel alignment angles were enough, and I guess it's his job to calculate the Mz contribution of each change. We bought a different rack with a slower ratio, (not because it was a slower ratio, but because it was gold in colour and price). however I find later the suspension guy had actually decreased the toe-base significantly (from 70mm to ~63mm) to match the rack so we maintain the same overall steering ratio. It seams fair, but not surprisingly we had quick steering with similar feel to last year. Some spacer plates where made to increase the toe-base by 15mm which worked very well and the car is drivable, however people still comment the steering is quick with high effort.

I had planned to make a new steering wheel that is 5mm larger in radius ( I agree with incremental changes and keeping things simple). We made the steering wheel but a local business that was to trim the wheel went out of business and out steering wheel went in trash with the rest of the company. So we used last years steering wheel.

A solution now would be to remake our spacer plates to again increase toe-base further. We also have a mild steering lock issue. The inner rim just about touches the wishbones and we struggle to get around some cones.

[Z]

Jonny,


After talking to another team I chose some very low numbers for KPI, caster and scrub,...

I sent a PM to Adam about the heavy steering issue.

But for now can you give your numbers for,
1. KPI (aka SAI = Steering-Angle-Inclination, from vertical, in end-view),
2. Castor (angle in side-view),
3. Offset (aka "scrub radius" = end-view distance between SAI-ground-intersection and nominal-centre-of-wheelprint),
4. Trail (side-view distance ...),
5. Degrees of full hand-wheel movement (lock-to-lock),
6. Corresponding degrees of front-wheel movement (lock-to-lock).

By FSAE standards you have more than average weight on the front wheels, plus biggish sticky tyres, but once the car is rolling a good steering geometry should give very light steering feel.

Z

[BillCobb]

A useful procedure to employ is called a "Friction Breakdown Test". It can be very revealing. Use a torque wrench on the steering wheel to measure the scrub torque with the car stationary. Then place the steered wheels on grease plates and measure once more. Then disconnect the tierods and measure again. Finally disconnect the column from the rack and get a value.

You maybe quite surprised to find some unexpected steering loads coming from ball joints, rack internals, swivel joints, u joints, whatever, under actual loading conditions. The grease plate test is where the eyeballs usually start coming out. "Ain't supposed to be that way" is a commonly heard expression. If your rack is tight, the rack pushaway mechanism (to prevent tooth separation) can be a problem, but if you modify it there is a chance you will jump a tooth during fast turning maneuvers.

There's always air pressure. If the cause of high effort is your tire MZ output, chances are you also have a problem with compliance understeer in the front end if your restraint structure is undersized. The car changes character when the MZ runs out at high cornering levels and becomes bi-polar. That's usually not a preferred driving situation.

[MCoach]

Jonny,



I sent a PM to Adam about the heavy steering issue.

But for now can you give your numbers for,
1. KPI (aka SAI = Steering-Angle-Inclination, from vertical, in end-view),
2. Castor (angle in side-view),
3. Offset (aka "scrub radius" = end-view distance between SAI-ground-intersection and nominal-centre-of-wheelprint),
4. Trail (side-view distance ...),
5. Degrees of full hand-wheel movement (lock-to-lock),
6. Corresponding degrees of front-wheel movement (lock-to-lock).

By FSAE standards you have more than average weight on the front wheels, plus biggish sticky tyres, but once the car is rolling a good steering geometry should give very light steering feel.

Z

Don't need to give away everything.

Johnny,

What's the steering ratio of the current car? As in (front tire degrees) / (steering wheel degrees) ?

[Westly]

Billcobb is correct that friction can play a large part in steering effort - we had a significant increase in our steering effort purely due to how we staked our spherical bearings in the a-arms increasing the friction.

I would also recommend physically testing any of these changes - it can be quite easy to allow for a lot of these setting (kpi, castor, scrub, trail) to be adjustable during design. One thing we were told by some "experienced FSAE people" was that our scrub was too big and this was the cause of our heavy steering. This was easy to test with rebuilt wheels with different wheel shells increasing our scrub radius from ~60mm scrub to ~100mm. The outcome - it had no noticeable effect in steering effort. So ensure you test yourself!.

We found that significantly reducing our mechanical trail (down to 0mm - which i'm sure ill get told is too low...) and slightly reducing the steering rate had the biggest impact on reducing our steering effort (and also significantly increased feedback to the driver). This corresponded to a ~1 second a lap improvement over a track ~60seconds long. This small change was likely one of the biggest improvements in the design of the car around those years (was much more significant than the ~15kg of weight dropped that year) - But don't take my word for it test it for yourself.

Our settings back then were:
1. KPI = 0
2. Castor = 4.5deg
3. Offset/Scrub radius = 60mm
4. Trail (side-view distance ...) = 0mm (was adjustable 0mm to 40mm)
5. Degrees of full hand-wheel movement (lock-to-lock) = 200deg
6. Corresponding degrees of front-wheel movement (lock-to-lock) = cant remember this one from memory - was adjusted until we could easily make the minimum radius hairpin as per rules

[BillCobb]

Lock to lock references are irrelevant to the effort recipe. What we/you want to know is what the effective steer arm length is and what either the pinion radius or the pitman arm length is. If this is a State secret, post it on Hillary's server. On grease plates, you will be able to feel/measure the effect of rear suspension roll stiffness, especially with that much caster. That's not a criticism of your caster setting, for some tires, that's about right. The question is what the max steering wheel rim force your driver feels comfortable with and would you dare run a 300 mm rim diameter to get it (for example)?

[Z]

MCoach,

Given their "heavy steering", I don't think UTAS has any comp-winning steering-secrets to keep!
~o0o~

Westly,

Although it is feasible to adjust Trail independently of Castor, it is not common.

Did your system allow this? Could you keep Castor fixed at 4.5 degrees, while varying Trail 0 - 40 mm?
~o0o~

Bill,

I agree that friction could be a contributing cause to their heavy steering, but I imagine Jonny and Co would have noticed that in their more than a year of struggling with it.

Yes, "effective steer arm length [to] pinion radius" ratio is the important number to know, but that number varies a lot over the full range of steer-angles, so requires a curve to show the full range of ratios. Just knowing the two lock-to-lock numbers gives a good average ratio to start from.

For example, if they had a total of 100 degrees of hand-wheel movement, AND they could also get around the hairpins (though I think they struggled there), then I would suggest they lower the ratio (ie. try 200+ degrees of hand-wheel rotation for same amount of front-wheel-steer).
~o0o~

Anyway, I didn't get a close look at their geometry at comp, but I reckon the problem is a down to some of the above numbers.

Z

[BillCobb]

OK, that makes sense to me, I'm not used to counting lock to lock because its just a rack length on a wide car with a specific wheelhouse size for a required turn radius a lot bigger that for these cars.

I suggest an additional steer effort puzzle piece: screwed up (by design) or by careless assembly of U-joints in the steering column and intermediate shaft.

It only takes a few mm of mis-position to create these effort changers. One condition might actually help if you can accept the reduction in gs by steering gain. Determining these geometricly induced ratio changes is just simple trig analysis of co-linear input and output shafts via Excel. A single u-jointed steering part actually has a lot of potential to address the steering gain and effort requirement/goal. There are quite a few production vehicles who use this technique instead of springing for an expensive nonlinear R/P gear.

[Tim.Wright]

Was that asymmetric one from the far east by any chance?

Or west from your reference frame..

[BillCobb]

The pix I posted are from one of my simulations which accepts the coefficients for a ratio function usually determined from measurements of test cars. The function acknowledges ratio variation due to bad i-shaft geometry, bowed racks having designed in ratio nonlinearity, and asymmetry due to steer arm height and length choices or oversights. Test output data channels then become obfuscated by the nonlinear steering control.

A summary from thousands of such tests reveals manufacturers who don't care, some who are anal about it, and some who we don't know about because somebody took the steering system apart to run cost and content studies on it and didn't have the proper tools or savvy to put the stuff back together correctly.

Some Euro-sleds make use of a deliberate INVERSE valley (which corresponds to a steering torque valley) and others which attempt to address an on center understeer softness by using a valley ratio to increase the gain as you corner more. When there was an understeer test result with bizarre asymmetry, first thing to suspect would be a soft tire, or a bad suspension bushing or a wheel alignment problem. Not finding anything wrong, I started measuring ratio from a differential of the steering wheel to road wheels function. Then, correcting the understeer calculation for a new and improved ackermann gradient produced a perfectly wonderful looking symmetric set of cornering compliances and understeer. Correcting the steering parts then produced a much different feeling car, too. It didn't take long for the responsible Development Engineer to learn the relationship between how the car felt and what the likely steer ratio issues were. That's what's expected in the Profession.

But, back to the message: a FSAE car with one Cardan type joint in the upper steering system can acquire quite a range of gain or effort attenuation if the design is implemented on purpose with forethought.

[BillCobb]

See if you can get this working in Matlab. I've punched up a GUI deal to play the graphics. Run the '.fig' file. This generalized function can pretty much deal with the entire world of good and evil steering ratios. Credits to my buddy Keith Adams for dreaming this up. If any of you Rembrants come across some data for your car, try it out ! Use "lsqcurvefit" to punch out the coefficients.

[Z]

But, back to the message: a FSAE car with one Cardan type joint in the upper steering system can acquire quite a range of gain or effort attenuation if the design is implemented on purpose with forethought.

Yes, indeed. And I recall UTAS having two UJs in there...

Given the many pages spent discussing UJs on the UTAS-14 thread, I hope they got this right. But if not, then easy fix.

Just re-clock the UJs!

(May require new hole drilled at 90 degrees to existing hole...)

Z

[Jonny Rochester]

All this talk of steering on the UTAS thread! The car, as I drove in endurance, wasn't all that bad. There is reduced lock (hard to miss all cones in tight bits) and the steering is a bit quick, but not disastrous. (Quick steering confirmed from feedback on the driver swap day).

This maybe a sore point for the team because some members are now against me, saying I have said wrong things or misrepresented the team on this forum. I'm not changing any of my comments, as my stories are my own account to the best of my knowledge, and if I quote a measurement from our car, it is often directly after measuring it with a ruler myself, even if an approximation is needed due to an awkward angle. I was "ergo" leader at the start of the year but other people followed that up as I became electrical leader.

While different uni joint angles maybe interesting, I don't think it is directly applicable because we have a double cardan joint which I believe equals a CV joint. I/we tried to get an equal angle on each uni joint, however when looking from above the 2 joints are misaligned by about 2mm. I think I have the basics covered as far as looking for friction in different joints etc. When I place a rag under each front wheel the steering is significantly lighter. And it is not hard to calculate a ratio... Regardless of what the exact design was, I can now photograph and measure our actual car, as it was set for endurance. For your information...

Steering wheel lock: 180 degrees total
Steering rack travel: 56mm total (limited with spacers so rims don't hit wishbones).
Toe-base: about 83mm +/- (I measure 63mm original plus 20mm spacer, this is disputed as I'm told I don't understand where to measure).
Steering tyre angle: as pictured (not quite enough).

929930931932933

You may see the steering column bearing is plain bearings instead of cartridge ball bearings, with a stainless steel column in IGUS plastic bushes, something I tried this year and it seamed to work well except for just a small amount of movement.

[BillCobb]

A double Cardan mechanism CAN be a CV system IF the shaft inputs and outputs are colinear and the shaft's input and output angles are equal. If the input and output shafts are NOT co-planar, (as in just about every FWD car ever made), then a phase angle is required on the inner shaft to correct the phasing. A spreadsheet made up to analyze this geometry is a useful design tool. It will tell you that the greater the shaft angles are, the more sensitive the design is to a misbuild situation that greatly affects your steering feel and gain characteristics. At 34 degrees, a 1 mm column mount error will induce a detectable 'lumpy steering' feeling. It will drive you nuts if the effect is at the zero steer position because it manifests itself as a steering lead in the car and the effort sense will be different left and right.

Steering intermediate shafts (as they are refered to in the industry) seldom have input and output shaft angles greater than 32 degrees for reasons you may have to experienced for yourself. Forces within the U-joint cross caps can get so large they will pop out the bearing rollers or end caps. I have a 2010 Vibe right now which has popped a cap on the u-joint at the EPS connection.

A way to tell if your own personnal FW Driver has a problem is to count the threads showing on each tie rod. They should be exactly equal and the steering wheel should be centered when driving down a straight flat road. If not, Houston, we have a problem. An alignment shop or dealer will centered the steering wheel by adjusting the length of each tie rod to get the toe and the steering wheel looking OK. But, this 'trick' will induce 'orbital steer' or other leads and pulls in the vehicle because the rack is not centered, the ride steer is not symmetric and the brake steer isn't either.

Some candidate steering CV joints are available for quads, UTV, ATV, and MTV viewers if your packaging department runs into a jamb.

I almost forgot to mention that the torsional stiffness of your steering shaft goes down dramatically as you increase the joint angles, too. That adds compliance, hysteresis and a loss of road feel. Another downside to FWD. [Friends With Donuts].

[CWA]

...There is reduced lock (hard to miss all cones in tight bits) and the steering is a bit quick, but not disastrous. (Quick steering confirmed from feedback on the driver swap day)...

Hi Jonny, any chance you could elaborate on your description of 'quick steering' a bit?

I only find this comment interesting because in another thread here (the 'pitman arm' thread) we discussed the negative implications, if any, of having too much steering gain (vehicle response output vs steer displacement input). I was quite convinced by the arguments put forward that high response to displacement as a control input is not necessarily bad (Z's fighter jet, Charles's kart), but this insight also went against what I learnt during my FSAE experience (and in industry, as BillC will testify).

I wonder why you and I might have found that in an FSAE application it is possible to have steering that is 'too sensitive' whereas in other applications this may be an acceptable, desirable, even necessary attribute.

P.S. Merry Christmas all

[StefStam]

IMO its better building a reliable and yet simple car for such competitions. Most teams in Europe talk about judges pointing out too complex mechanisms in FSAE car. Although myself i want the team to build the car with full aero packge,its kinda difficult though
Keep in mind that you need to keep weight as low as possible