Tube discussion

[Kenny T Cornett]

I've been thinking a lot about tubing recently and wanted to ask a few questions to some of you all frame builders to see if I could get a sample of info.


The question stewing in my brain is about chassis weight due to tube wall thickness. Aside from the stipulated tube parts, what sort of wall thickness are you all running.

More specifically, what kind of linear amount of each wall thickness are you using.

I know that frames vary drastically from team to team, but I'm looking to set up an "average" tubing amount and then play with weight figures based on tubing wall.

The whole idea for this is that our advisor is pushing materials simplicity and homologation (i.e. as few tubing sizes as possible).

My interest was peaked when a team mentioned that they were using 1" tubing for their control arms and when I showed our advisor he wanted us to develop a mini-study to determine the feasibility of using a single tube size (other than roll hoop tube).

Thoughts? (sorry for the long windedness)

[Kenny T Cornett]

I've been thinking a lot about tubing recently and wanted to ask a few questions to some of you all frame builders to see if I could get a sample of info.


The question stewing in my brain is about chassis weight due to tube wall thickness. Aside from the stipulated tube parts, what sort of wall thickness are you all running.

More specifically, what kind of linear amount of each wall thickness are you using.

I know that frames vary drastically from team to team, but I'm looking to set up an "average" tubing amount and then play with weight figures based on tubing wall.

The whole idea for this is that our advisor is pushing materials simplicity and homologation (i.e. as few tubing sizes as possible).

My interest was peaked when a team mentioned that they were using 1" tubing for their control arms and when I showed our advisor he wanted us to develop a mini-study to determine the feasibility of using a single tube size (other than roll hoop tube).

Thoughts? (sorry for the long windedness)

[Jersey Tom]

If its triangulated well, I'd go as thin as possible. If its structural and in bending (like engine mounts sometimes are), it should be much heavier section. Especially if its long.

Its a racecar. Engineer stuff to be as light as possible. Perhaps personal preference but I would definately not use 1" control arms nor a single wall thickness. I can't imagine the price break on that much more steel of one size would offset the weight. Even for a 1st year team.. chassis and suspension steel should only be around a grand.

[Mike Cook]

Your advisor sounds like he doesn't like to win. Besides the standard tubes you should run 1" 028 everywhere. Should be able to get the chassis about 1500+ ft-lb/deg.

A arms should be made of .5x028 or .625x028.

[Sgt Slaughter]

I agree with the advisor on several points. Yes it may not be light for you to use 1" .065 tubing for every part of the car, but a first year team is not out to win anyways (I'll probably catch some flack for that statement). Overbuild as a first year team for safety if nothing else. Teams have rolled a arms over in the past because they were trying to cut too much, and nobody wants to not finish at comp because of that. Lets also remember one thing here guys, the ounce you save by using a .028 tube is the ounce you're gonna need when you hit something by accident. Lets look at it from a manufacturing standpoint. If you use CNC bending, it is easier and cheaper to stick with the same sizes and bend radii for as much as you can. Similar material types produces less waste and will give an inexperienced welder less grief than an .028 wall. First year teams have to look at getting a car done before worrying about the extra 5 lbs of frame weight. Just my .02

[ScottW]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I agree with the advisor on several points. Yes it may not be light for you to use 1" .065 tubing for every part of the car, but a first year team is not out to win anyways (I'll probably catch some flack for that statement). Overbuild as a first year team for safety if nothing else. Teams have rolled a arms over in the past because they were trying to cut too much, and nobody wants to not finish at comp because of that. Lets also remember one thing here guys, the ounce you save by using a .028 tube is the ounce you're gonna need when you hit something by accident. Lets look at it from a manufacturing standpoint. If you use CNC bending, it is easier and cheaper to stick with the same sizes and bend radii for as much as you can. Similar material types produces less waste and will give an inexperienced welder less grief than an .028 wall. First year teams have to look at getting a car done before worrying about the extra 5 lbs of frame weight. Just my .02 </div></BLOCKQUOTE>

I agree with most of that but 1" A-Arms are more than overkill and could probably cause more problems in fitment than they solve in cost/strength.

[B Hise]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Lets look at it from a manufacturing standpoint. If you use CNC bending </div></BLOCKQUOTE>

so... lets keep it simple and get some CNC machines. wrong.

Its not hard to do some statics and figure out the loads on the a-arms. Being that this is an engineering (and management) competition, the first year teams should challenge themselves. There is enough information floating around for a first year team to build something decent if they keep their heads on and work hard. Yes I know how hard it is, my first year of FSAE was on a first year team and it sucked. But we worked our asses off to establish the program.

In closing to this rambling. I think getting one diameter of tube is insane. Draft up some wireframes, do some simple analysis, and then figure out what you need. I'll see if I can dig up our 07 steel order in the meantime.

-B

[formula_wally]

Sample wall thickness used on 1 inch tube frame, highly-rigid light frame.
0.028
0.035
0.045
0.065


1 inch control arms are usufully only if your first years are dumb enough to try and lift the car by the upper eh's.

Get an advisor that wants to win, and not save money&lt;:

[Frank]

Triangulation (efficient load paths) is the key to a stiff space frame chassis, BUT BUT BUT....

Even the best chassis will still have ALL members in bending.

Think about it. A long thin member (tube) will encounter deflection in BENDING, not in the axial direction.

Therefore, you want each member to be efficient.

Efficient sections are thin wall.

Apart from the mandated areas for impact protection, you want to use the thinnest tube PRACTICABLE.

We use 035 in most areas, because we have trouble welding 028. But you may find welding 028 easy.

For us the tubes used are
5/8*035
3/4*035
7/8*035
1*035

If you can weld it, then use
5/8*028
3/4*028
7/8*028
1*028


I'm not up on the rules these days but we also used a slightly stiffer (with comparable cross sectional area) tube for the roll hoops. It was 1_1/8*083.

5/8*035 or 5/8*028 is common for suspension links

1/2OD or 9/16OD might be OK for short(ish) links

Once you hit a few things, you might appreciate 5/8OD, but we all try to avoid these "incidents" http://fsae.com/groupee_common/emoticons/icon_smile.gif

Do some calcs (buckling calcs) and FEA (dont mesh tubes, use a node/beam model), fiddle with some tube sizes, and note the overall efficiency of the structure in moment/angular deflection/mass of structure and make up your mind about what mix of tube you'll need.

Hope this helps

Frank

[Jersey Tom]

TIG'ing .035 or .028 isn't too bad with a little practice... low amps, shade 9 helmet, .045 filler. But hot damn don't try welding .035 or .028 to a .095 tube or somethin like that!

[Mike Cook]

An additional few things.

Welding 028 is not easy. I don't care what anyone says. The notches have to be perfect and I don't have much faith in the welds. For frames they are usually OK.

It shouldn't be difficult to drive to a steal tube supplier and get different wall thickness tubes.

It shouldn't be difficult to order different wall thickness tubes.

Call dillsburg, PA - its an old rickety shed complex that has steel out the ass. Don't get the guy going about college students. Don't tell him how to live. Don't say anything. Don't change you order. Get there, load the steel and leave.

FEA is cool if you know your load cases. However I never really learned much from FEA'ing a frame other than what was already obvious if I had sat down and stared at the frame.

Instead of doing a tube study, build a car.

Calculate your loads at the CP. Assume something like 1.5 g's. Assume a 700lb car. Calculate the load this applies to your spindle, hub upright and finally ball joints. calculate loads in all the members. take these loads and apply them to the frame. Think about what might happen. Triangulation is your friend. Nodes should have three members. Don't have any steps in the frame.

Enough

[Sgt Slaughter]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> so... lets keep it simple and get some CNC machines. wrong </div></BLOCKQUOTE>

And a first year team should be opposed to CNC bending why? We've bent cromo by hand before on a manual bender and it blows. It was completely worth it to have someone bend tubes to spec instead of screwing around bending your own tubes. As far as the complexity in having it CNC'd, I can't see where it is, you need a print no matter what you do. Consistent bend radii are not difficult to do by any means either. I admit that packaging a 1" OD a arm and control links would be difficult, but hey what on this project isn't.

The point I think we're still missing is the fact that they need to finish a car. Let's face it, if they're still looking at wall thicknesses they're already behind. Michael Royce said at the Cedarville seminar that race day comes no matter what and that a bad car that makes it there, is better than the best car that never does.

Again, just my .02

[Mike Cook]

Can you please let the rest of the world know where you store your magic closet of CNC tube benders because I'd like to know. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Sgt Slaughter]

Troy Tube bends for us

[rjwoods77]

Just a comment on the "this is a rce car so it should be designed just enough" I think is good in theory but the point of the competition is to design a car for the end user which is going to be some idiot person who does idiot things. What are you to tell the idiot you drops the nose of the car by accident when unloading and bends or breaks any peice of equipment on one of these cars. "Sorry sir but these cars were designed to be finicky pieces of equipment that are meant to be handled with kit gloves. We didnt design it for the off chance you might break something. That will be 400 bucks for the carbon nose cone and 150 bucks for the a-arm." I remember one of the design judges saying at the seminar a colin chapman statement about weight. While he is totally correct for a pure race car, I think he is incorrect for saying this about a car meant to operate in an unsanitary, non organized parking lot and not a Formula 1 paddock.

You can get shops to cnc bend tubing for about 12 bucks a bend and they will do it perfect and do it right the first time. Typical fsae students will blow this on wasted material alone. We had two full chassis bent for 500 bucks and they were dead on. We got half off so a single chassis (3-4 bend roll hoops,2-2 bend rear support hoops and 2-1 bend lower chassis rails) cost 500 bucks. There was almost no waste and they came out perfect and I mean perfect.

Another thing to consider is your joint sizes. We are using aurora hab-4tg joints everywhere on the car. Either using a rod end or a spherical holder you will need a 1" diamter to hold the spherical. Well weld the holder into a 1" thin wall tube and the holder is the same width as the tube. Yeah it is a littel overkill but the added weight is really marginal, it looks better, it is much more rigid and solid of a design. A dumbass customer will never hurt these short of talking a hammer to them. More later. Got to go to lab.

[rjwoods77]

Just to add.

1 foot of 1"x0.035"= .359 lbs
1 foot of .625"x0.035"= .22 lbs

Lets assume that there is 3 feet of tubing per a-arm.

On a typical 8 a-arm car you will save 3.336 lbs(.417 per a arm). The longer or shorter the less or more weight obviously. That is a lot of weight for a seasoned competitor but nothing to cry about for a new team intersted in cost savings and getting things done with a minimum or reasources. The 1" stuff will be way stronger than the 5/8x 0.035 tubing that is known to be strong enough if you dont hit anything hard with them, dont ratchet strap over them, etc. Thats also assuming that the 5/8x0.035 will work for what you are looking for in terms of stiffness, deflection,etc.

[B Hise]

JD2 tube bender w/ 3" x 180 deg die and shipping = $500. With some care and sense, you can bend your tubes dead nuts. So with the $500 youll save the next years, you can buy tires (3 anyway) to test and learn something. You also never have to wait on anyone to make your parts. Autonomy is key.


Getting into the gray area of the comp, yes these cars are supposed to be built for customers, but I dont see the judges giving out points for the added weight and durability. I could be wrong, but if thats the case then I'm adding a cow catcher for the enduro. The best way to not wreck your car is to not let monkeys unload it, run it into and/or over things, and do generally dumb things like lift the car by the a-arms. I'm well aware that things happen, but I'm here to learn how to design racecars. There is no way I would design allowance for dullards to wad my car up.

- B

[rjwoods77]

I hear what you are saying about the judges but you have to convince them of your design direction. I think most just go with the "but the judges wont like it" on base concepts because most people dont know how to explain things in a persuasive manor. The judges always tell us. Do what you want but you better have a good reason. Funny story. I am from Rochester. I stopped over at RIT to check out what they were up to last year. I went to go sit in their car. I had a kid give me a dissertation on what not to step on, how to shimmy into the car, i couldnt wear my sneakers, and a couple other things about this is fragile and that is sensitive. I thought to myself right there that this is the furthest thing from any autocross car. Might as well been trying to drive a glass palace. Nice car though. Not a diss on their engineering. Just it wasnt designed for the intent of the competition. The kind of people that would drop dime for a car like this would be the Joey Bag-O-Donuts crowd (at least in New York and yes I am Italian) who like to drop cash to have fun with neat toys. The Lake Ontario Italian navy is a perfect example.

[Dan Nauts]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kenny T Cornett:
Aside from the stipulated tube parts, what sort of wall thickness are you all running.

More specifically, what kind of linear amount of each wall thickness are you using.

The whole idea for this is that our advisor is pushing materials simplicity and homologation (i.e. as few tubing sizes as possible).

My interest was peaked when a team mentioned that they were using 1" tubing for their control arms and when I showed our advisor he wanted us to develop a mini-study to determine the feasibility of using a single tube size (other than roll hoop tube).

Thoughts? (sorry for the long windedness) </div></BLOCKQUOTE>

As a first year team, we used 1" 0.065 everywhere. It was easy to build, easy to order material, and one less thing to worry about.

As a second year team, most everything that's not required to be 1" OD, isn't in bending, and suspension doesn't attach to it, is 5/8th's 0.058. Yes we could probably get more weight off the car, but we're still focusing our attention on other systems of the car.

Making the common tube size choice has made fab and tooling real easy, as well as ordering and storage (every piece of 5/8th's in the shop is 0.058).

Your advisor has a very good point. Homologation is a damned fine goal; common tube sizing, common hardware, common brackets and tabs,... all will streamline the design and fab of the car.

And as always, "pick your battles", if you need to spend a significant amount of time trying to minimize tubing, and increase stiffness, but gain minimal returns, you probably should have focused your effort elsewhere. Does not always hold true, but unless your an established team with proven design ability, where design doesn't deviate much from year to year, time is best spent in areas with the chance for the greatest return on investment of resources.

[apoorav1989]

hello guys ..we are dng fsae for the first time ...so can anybody suggest what double wishbone A arms diameter and wall thickness should we take... ?

[RacingManiac]

Read the above....and if you really want just an answer, 5/8" .035, .028, and 1/2" .028....

[cwojcik]

Rob, I understand what you're saying about RIT's car not being for the casual autocrosser. But, there is a market for anything. While your car might be aimed at regular Joe who wants so store the car in a shed next to his lawn equipment, another car (RIT's for example) might be designed for the guy who really wants to win, and is willing to deal with an expensive, finicky, delicate car in order to get the best performance he can.