Suspension Analysis

[Robert Skapof]

Hello,
Is it possible to do analysis on suspension geometry without having a program like Adams?

I I have gone through SLA iterations and would like to take my parameters and insert them into a program like mathcad or matlab with textbook formulas.

[Robert Skapof]

Hello,
Is it possible to do analysis on suspension geometry without having a program like Adams?

I I have gone through SLA iterations and would like to take my parameters and insert them into a program like mathcad or matlab with textbook formulas.

[Zac]

Sure. What kind of analysis do you want to do? Dynamic simulation? simple kinematic analysis? matlab is a pretty powerful tool when you have a clear understanding of what you're trying to do.

Also, don't forget about the compliance your suspension is going to have once you actually build the thing. Spending hours trying to sort out that last mm of roll center migration isn't going to accomplish anything if your outside tire flops over positive during hard cornering.

[Mike Macie]

Absolutely...

You really don't need much to design a good suspension. Just some good vehicle dynamics knowledge and knowing what you want the car to do through different modes. Remember, testing is what makes a good suspension great.

[Robert Skapof]

Right now i have my hard points for the front and rear double wishbone suspension. I wanted know what kind of analysis i should be doing? I have RCVD. I want to know if the formulas in the book can be applied when the center of gravity is shifting due to body roll and pitch.

I guess the better question is how do i solve for the CG location under roll and pitch? How much force will it take to make the body roll and pitch given my suspension set up?

This is the type of analysis I want to do.

[scott_rfr]

Just take your cg to be stationary. In RCVD there is a whole section about calculating your roll gradients. Find that section and make yourself a nice spreadsheet. You should be able to input mass, cg location and height, tire spring rate, wheel rate, rch, pch, arbs. Then apply a lateral or long. force and you can see how much weight transfer you get at each axle, what your roll gradient is etc...

As for kinematic analysis you need to determine what you want your camber curves too look like, rc height, steering feedback etc. Then look at how each one changes with some motion. Then was that change good, bad, or acceptable?

Scott

[J.R.]

I wouldn't worry too much about lateral RC migration, just so long as the vertical displacement is relatively small. Search for the Roll Center Myths paper on here that was posted a few weeks ago.

Do you have tyre data? This will tell you how you want your camber curves to look, and give you the steering feedback etc. Without it, it's all guesswork.

[Robert Skapof]

i do not have tire data but i am planning on using the hoosier R125A

[J.R.]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Robert Skapof:
i do not have tire data but i am planning on using the hoosier R125A </div></BLOCKQUOTE>

They don't make the R25A (think thats what you meant) anymore, it's all R25B now. Basically same compound, just supposed to last longer with no performance decrease, or so they say.

Can't really share any info about designing around the tyre's, because of the confidentiality agreement, but if your team has $500 to spare, it's the best investment that you can make. There are even some codes floating around these forums to make the stuff really easy to see.

If you are in the analysis phase of design right now, is your design for 2k9 or 2k10? Either way, try and find some sponsor to pay for the tyre data, at FSU, you should have some good Alumni that you can turn to! http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Wes Johnson]

Just wanted to add that a good start for your kinematics analysis could just be a simple CAD program. You can get alot done with just a 2d sketch.

-WJ

[apoorav1989]

actually we are doing fsae for the first time ..so we want help..i want to know that what all should we start reading in suspension system and that too in appropriate amount...i will be really obliged if u help me

[J.R.]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by apoorav1989:
actually we are doing fsae for the first time ..so we want help..i want to know that what all should we start reading in suspension system and that too in appropriate amount...i will be really obliged if u help me </div></BLOCKQUOTE>

That depends on how in depth you want to go. Start with "Tune to Win" by Smith, and read the suspension, springs and dampers section several times until you understand what he is saying. More advanced stuff in "Race Car Vehicle Dynamics" By Milliken and Milliken.

I started on the suspension team the day before design at Detroit last year, and I haven't stopped reading since... That being said you can design an "adequate" suspension on basic principles. To design a "good" suspension takes a ton of understanding, and tyre data....

That being said, a first year car should be more concerned about making sure all the bolts fit in the correct places, so start with "Tune to Win" for this car and go to "RCVD" for your next cycle. Good Luck!

[Robert Skapof]

Can someone give a quick example of designing suspension based on the tires (tyres). I am going through this suspension design, keep in mind we do not have a frame, and with the calculations to get something of interest we dot have a CG location and it all would be guess work until we have something built or in solidworks the way we plan on building it.

[J. Vinella]

Robert,

As for your CG I would poke around and see what other people have gotten once their car is built, if you can find that. Take a most conservative estimate that is all you can do at this point. Your model will be close if it is accurate, but as you said it is all guess work and you will need to be satisfied with that.

As for an example on "designing (a) suspension based on the tire": Using tire data, what position can you get the most out of the tire in cornering? What position can you get the most out of the tire in braking and forward acceleration? How much of your time in this competition is spent cornering and how much under braking and forward acceleration? Answer these questions. Then compromise or invent some crazy new suspension geometry that gets the best out all the modes. Then reflect on your tire data...wondering if it is telling you the real story. Suspension is extremely complex and for a first year team I would just design something that can be adjusted greatly. One could wright an entire book on vehicle dynamics and suspension systems and people have. Many good books are referenced above. Just pick something and go with it, you will learn so much in testing and you want to leave yourself time for that.

Lastly, I have heard and read many say they are designing their suspension system for the tires. That is missing the whole picture. I say design your suspension for this competition, the tires are just the most important variable. Take if from me that is a pearl of wisdom. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Robert Skapof]

John "Jack",
Thanks for the advice. I guess I am off to buy some tire data from hoosier, along with some tires of course. Another thing that seems to be getting to me is the spindle design with the wheels, hub, rotor, bearing, and uprights. The hub is made or bought?

[Hector]

Tire data isn't something you can really buy from Hoosier. It's done by independent research groups and generally costs a lot of money. Since teams cannot afford it individually, the FSAE Tire Test Consortium was formed. Teams pay $500 to join the consortium, and you get all data on tires tested. You can read more at

http://fsae.com/eve/forums/a/t...607348/m/15510029041 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/15510029041)

This link is also stickied at the top of the General Discussion page.

EDIT: This is probably the best $500 you can spend. Not only will you have more info to design suspension, drivetrain, and braking around, but good luck talking to a suspension judge at competition if you don't have tire data.

[Eklavya Singh]

I mailed Hoosier to get a quotation for their tires. They sent me the tire data along with that. I'm sure that's a dependable source without joining the consortium. You could try that.

Eklavya Singh
Technical Head
Formula Manipal

[Robert Skapof]

was the tire data in depth? or basic?

[apoorav1989]

what is meant by tire data....and where would i get that http://fsae.com/groupee_common/emoticons/icon_confused.gif

[Robert Skapof]

You can get the tire data from the places listed in this thread. Tire data can give you lateral force generated at the contact patch of the tire based on slip angle, normal force, temperature, tire pressure, road conditions, camber, etc.

If you can maximize the lateral force at each tire under all conditions you have built yourself quite a car, which means putting each tire in the situation to generate the maximum lateral force.

I am only learning this as we speak so you may want someone else to answer these questions or go through these forums. I believe you can find answers spread out through the entire forum.

[Anvit Garg]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Eklavya Singh:
I mailed Hoosier to get a quotation for their tires. They sent me the tire data along with that. I'm sure that's a dependable source without joining the consortium. You could try that.

Eklavya Singh
Technical Head
Formula Manipal </div></BLOCKQUOTE>

Did they just send you the spring rates posted on their website or actual raw/processed data regarding aligning torques and lateral/longitudinal forces?

[apoorav1989]

what is meant by track change and track width ...?

[Eklavya Singh]

@ Anvit
@ Robert

They sent me the stuff on their website. That's only the spring rates. If nothing else it'll at least get you started on the ARB calculations and also ride rates.

Here's the link
https://www.hoosiertire.com/sp...0%20x%206.0%20-%2013 (https://www.hoosiertire.com/spring.htm#TIRE%20SIZE:%2020.0%20x%206.0%20-%2013)

Eklavya Singh
Technical Head
Formula Manipal

[Robert Skapof]

Problems with design!

doing a 23 mph, 55 ft, radius turn. Pulling .64 g's and I will get a front lateral weight transfer of 91.4 lb and a rear of 161 lb. This means that the inside front tire has 43 lbs on it and the inside rear tire has 3 lbs on it.

To me this sounds terrible, though i have only gone through one iteration (followed along RCVD 16.2)

My wheelbase is 63", front track = 55", rear track = 54". I am thinking I need to make it wider but i feel this car cannot make a turn for its life and i am changing things around a lot right not to compensate but it is not getting much better.

[Zac]

check your units and input parameters.

[Robert Skapof]

Zac,
Thanks, it was my units. In one spot I forgot to divide by 12. The book does their calculations in feet and I have all of mine in inches. Usually i use mathcad but I am having licensing issues with matchcad 13.

Cheers,

Rob

[Silente]

I am interested to know how you would work on tires data to define suspension main aspects, above all i am interested to know how you would work to define your steering geometry on the base of tire data.

Can you give me any advice on this or an example abot the way you would analyze the problem?

I think we have to guess something about slip angles to work on that matter, also if you know corner radii or you already have an idea on how your Lateral Load Trasnfer is going to be because of bars and springs you choose... Isn't it?

[HenningO]

Silente,

One example of when tire data can come in handy is when you are determining what Ackermann your steering system should have. By looking at what slip angles the tire is the most efficient for a range of vertical loads, tire data can help you select Ackermann (ie. difference in inside and outside tire slip angle).

Your lateral load transfer will mainly depend on your lateral acceleration (which is a function of speed and corner radii). What bars and springs will tell you is the distribution of the lateral load transfer between the front and rear axles.

[Silente]

Hi Henning,

thanks for your reply.

Ackermann is exactly where i wanted to go..

But my doubt is a little bit different. Let's say you know exactly at which slip angle your tire give maximum cornering force for a given vertical load.

You would produce a steering geometry which only gives a dynamic toe out equal to the difference in the SA where you have your maximum cornering force at the normal loads you would expect on your tires?

How could you really evaluate SA in a certain radius corner?

[HenningO]

I'm not sure if I understand your first question, but if I interpret it as a statement, then you're right. With the "right" ackerman, both your inside and outside tire will be operating at the SA which generates the highest lateral cornering force for the given vertical load.

It should also be mentioned (and have been discussed previously on this forum) that in FSAE transients is where you'll get the winning edge. Running non optimal SA could be benefitial as it may be used to get your car faster in to the bends...

For a given curve and speed you can calculate the total tire forces that are required to keep the car on the track. For steady-state cornering the yaw rate need to be constant. So, you can figure out the distribution between front/rear tire forces.

And when you know what forces each tire need to exert, you can figure out what slip angles they are operating as. Just remember that the slip angle is equal on both tires of an axle - it makes calculating things a whole lot easier.

How close this method will get you to reality, I'm afraid I can't answer as I've never worked with SA sensors etc.

[oz_olly]

HenningO,

Above you said that they should remember that slip angles are the same across an axle. What do you mean by this statement? If you have a 4 wheeled vehicle travelling on a curved path, at any instant it will have a turn centre with a radius proportional to velocity^2 and lateral acceleration (radius=distance from turn centre to cg). The path that each wheel travels is tangential to to their respective turn radius (same turn centre as the vehicle) so unless you have rear steer desigined into your geometry you will have different slip angles across the rear axle.

It is quite easy to draw a sketch in your favourite cad program.

You also said that for steady state cornering the yaw rate needs to be zero. I agree this is true, however do you think it is easier to think that the yaw moment should be zero?

Cheers

Olly

[HenningO]

Olly,

You are right. http://fsae.com/groupee_common/emoticons/icon_smile.gif

I was cutting corners (no pun intended) when I wrote that about SA. To my defence however, I haven't done the math but something tells me the differences in SA on inner/outer tire is quite small. In the front the difference in slip angle is also depending on your Ackermann setup.

As for yaw rate/yaw moment I believe it's easier to first understand that the yaw rate needs to be constant (I'm assuming that's what you meant) and then see and understand the connection to the yaw moment and ultimately its relation to tire forces.

[Zac]

I haven't run the math for an ideal FSAE car, but I think the ratio of track width to minimum turn radius is small enough that you wouldn't see a huge difference in SA.

However, with a real car you're going to have some amount of toe change due to kinematic and compliance effects, not to mention your alignment settings, that will trump whatever difference in SA you see from pure geometry.

As a side note, I think the amount of compliance that a lot of these cars have in their suspension is ridiculous. I have K&C data on several cars. On the faster ones compliance accounts for a decent amount of what the suspension is doing but for a student built vehicle I think it kind of is what it is. With the slower cars it is another story. I've seen a couple cars where compliance dominates so much that the outside wheels will flop over into positive camber.

I don't know if teams aren't setting stiffness targets(for both components and installations) or what. But maybe they just need to design in some good old fashioned "road hugging weight?"

[PatClarke]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">With the "right" ackerman, both your inside and outside tire will be operating at the SA which generates the highest lateral cornering force for the given vertical load. </div></BLOCKQUOTE>

HenningO, This is not the only reason for selecting the 'right' ackermann!

Cheers
Pat

[HenningO]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by PatClarke:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">With the "right" ackerman, both your inside and outside tire will be operating at the SA which generates the highest lateral cornering force for the given vertical load. </div></BLOCKQUOTE>

HenningO, This is not the only reason for selecting the 'right' ackermann!

Cheers
Pat </div></BLOCKQUOTE>

Are you talking about running ackerman to generate a longitudinal force component to help generate the yaw moment? Or is there yet another reason?

[PatClarke]

Yes, among other things :-)
Pat

[Silente]

Again on this topic.

We probably understood that it is nearly impossible to state from blank paper what the right ackermann is, also if you have tire data.

But, if you know steering geometry, can you say what differences between left and right side slip angle you are going to have in a corner?

For example, if you have a good data acquisition system with speeds, accelerations, yaw rate, steering angle etc data, you can calculate a first value for a front SA and a Rear SA using bicycle model (Milliken), but how to know what slip angle you're going to have on left and on right side?

Is it possible to state such a thing in your opinion?

[SNasello]

Theoretically (asuming everything is rigid) you could determine the slip angle on each tire if you know where the instant centre of the corner is, the track width, the difference in inside and outside steer angles (with ackerman). This is the principle behind putting two slip angle sensors on the chassis (instead of one on each wheel). If you also know how much toe compliance you have you could get a better picture of what the slip angles are on each tire.

[PatClarke]

Stefan, you are not paying attention :-)
Do not forget that caster and KPI impart a vertical movement to the wheels as they steer!

Pat

[SNasello]

Of course, and to get the full picture you would also have to take into account the effects of roll steer and bump steer (if any).

[HenningO]

I don't follow here,

I admit that track width and wheelbase changes a tad during steering due to caster/KPI, which could affect the calculations Stefan mentioned two posts up. But vertical movement?

Pat, enlighten me! http://fsae.com/groupee_common/emoticons/icon_smile.gif

[PatClarke]

No sir, enough hints, you have to go away and think about it! =] =] =] =] =]

Pat

[Steve Yao]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by HenningO:
I don't follow here,

I admit that track width and wheelbase changes a tad during steering due to caster/KPI, which could affect the calculations Stefan mentioned two posts up. But vertical movement?

Pat, enlighten me! http://fsae.com/groupee_common/emoticons/icon_smile.gif </div></BLOCKQUOTE>

Vertical movement would change your normal load, and thus your "ideal" SA.

[PatClarke]

Steve is thinking along the right lines!

Pat

[oz_olly]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by PatClarke:
Steve is thinking along the right lines!

Pat </div></BLOCKQUOTE>

This is easy to see when you have the car on corner weight scales. When you turn the wheels away from straight ahead you will get a diagonal change in weight distribution. Try it and see for your self.

Cheers

Olly

[HenningO]

I guess I was thinking more in terms of geometry rather then normal loads.

What I came up I did with "my head", I haven't tried it in any software so margin of error is probably quite large http://fsae.com/groupee_common/emoticons/icon_wink.gif

Anyways, with zero caster and zero KPI the steering axis is going to be perpendicular to the ground. This means as the wheel steers the outer tie rod point is going to move in a plane that is parallel to the ground (no vertical motion of the tie rod point).

When we add caster/KPI the steering axis is no longer going to be perpendicular to the ground hence the outer tie rod steering point will move in plane that NOT is parallel to the ground. This means that the tie rod point will not only move in the X-Y direction but also in the Z direction.

It seems likely, so I am assuming that this change in Z will not be the same for the inside tie rod point. Therefore the tie rod link length have virtually changed, which is the same as the steering angle have changed. But since it will only take place when the wheel turns, I guess it's more appropriate to consider this as a steering nonlinearity.

Could it be this nonlinearity that we ought to take into account when we decide our Ackerman, Pat?

[PatClarke]

Well well =] =] =]

What do you think? Is it possible that the slip angle of the tyre will change with changes in the vertical load? Or that ackermann will cause differing vertical load changes as the tyres steer?

Now go back and re-calculate your ackermann...If you can!

Have fun =]

Pat

PS. Don't forget the loads across the rear axle will change too!

[PatClarke]

This subject isn't finished =] There is more!

Keep thinking

Pat

[apoorav1989]

what is the difference between anti roll bars and sway bars..and which one of them is better ?

[flavorPacket]

oh my

[Samo Simonian]

I would definitely use ARB's not sway bars, they're much more accurate. http://fsae.com/groupee_common/emoticons/icon_wink.gifhttp://fsae.com/groupee_common/emoticons/icon_wink.gifhttp://fsae.com/groupee_common/emoticons/icon_wink.gif


Sorry, they're the same for all I know, both are intended to give roll resistence in the form of a spring, that's all.



I gotta question:
Theoretical data is in my opinion a nice hint, but nothing more: steel flat tracks, belts or drums ain't the real dirty salty wet asphalt, and the guys from TTC for instance admit that for 10" Hoosier R25A the results obtained are relatively crappy.
What are people doing to counter this? What I did was just get an estimate and build in a shitload of adjustability.

Regards,

Samo