Chassis Stiffening, Suspension Ideas, Handling Improvements

[Inodoro Pereyra]

Onlytrueromeo: infinite weight is not a practical feature, it's a theoretical standpoint. The Universe doesn't have infinite weight, so, obviously, no car will.
You use that kind of statement to illustrate a point, or, like in this case, to isolate a variable.
For a suspension system of any kind to absorb all the energy applied to it, Newton's first law (action/reaction) can NOT APPLY, since the suspended body can not absorb/dissipate any energy. For that, it needs to be infinitely stiff (so it won't ABSORB energy) and infinitely heavy (so it won't DISSIPATE energy). That doesn't mean such a scenario has to exist. It's just a way to illustrate the extreme, so you can understand/quantify the middle ground.

Cars are not designed with flex for economical reasons. They are designed to perform under a given set of circumstances.
Increasing chassis stiffness can, depending on the method used, go a long way to improve the car's handling. It makes the car more predictable, tighter, more responsive, etc. But EVERYTHING in mechanics is a compromise, and so is this.
A stiffer chassis also makes the car more unforgiving, the ride harsher, and increase vibrations.
It's actually not so difficult to understand:
The car's suspension (what isolates the car from the road) is the shocks/springs, etc. The stiffer the suspension is, the more the car will "feel" the road's bumps and cracks.
The driver's suspension (what isolates YOU from the road), is THE WHOLE CAR, so the stiffer the car is, the more YOU will feel the road's bumps and cracks.

Simple: every bump transfers a given amount of energy to the car. According to the car's design, a percentage of that energy is absorbed by the suspension. The rest of it will be absorbed by the chassis, and dissipated onto YOU, and onto the atmosphere (the car will JUMP). ANY ENERGY the chassis can't absorb (because it's stiffer), will have to go somewhere else, and the only place it can go to is you, and the atmosphere.

So, depending on the method you use, you will have a varying degree of handling improvement, and a varying degree of ride quality loss.
For example: if you weld the doors shut, you will greatly increase your chassis stiffness, while only adding a few pounds to the car's overall weight, which means you can still probably use your standard shocks/springs without any noticeable increase in body roll, so the only difference you will notice in the ride quality is that due to the stiffer chassis.
On the other end of the spectrum, if you install a full on roll cage, you will be adding quite a lot of weight ABOVE the car's CG, which will tend to have a negative effect on your car's performance, both in terms of handling and power to weight ratio. So then you will have to go for stiffer shocks/springs, to keep body roll in check, which will make your ride quality go straight to hell.

[fowljesse]

I'm very interested in this topic, and looking forward to reading about how it works out!

Points to ponder:
From a metallurgists standpoint: What does't flex will break. It's called "shear strength". This is a gross oversimplification, but like the Zen master says "Be like the Reed, and flex with opposition". Something's got to give. This goes into the design of the car. People say "Why don't they make the whole plane out of the black box material?" These are just anecdotes for fodder. I believe chassis stiffening will help handling, but don't know at what price. Unless you're rally racing, I don't know how much it will help. I'm a freak about weight, since I used to race Mountain Bikes Semi-Pro, so I look at it that way. My car handles awesome (neutral). All the wheels let go at the same time, generally (no understeer). It's very predictible, and I love it. I know my MX-3 body is freaking stiff, because one time I was jacking it up just behind the front wheel, and the rear wheel came up, too. AND the door opened, and closed just fine. Every time I jack up the car, I test the door opening, and closing, and it's solid. Also, I drove 1 front wheel onto a ramp to do something, and it was balancing on just that wheel, and the opposite rear wheel, and the doors opened, and closed fine.

[onlytrueromeo]

Urgh I'm getting frustrated with all of you, because you're wrong, at least as far as I've learned in my multitude of physics/rigid body dynamics and engineering classes. I am not an automotive engineer, but I have done reading in automotive design books.

Jesse, yes, if something doesn't yield, it will fracture before showing signs of flexure. The "harder" something is, the less ductile it is, so it will tend to break..but this point has nothing to do w/ cars, since if our frames were brittle enough to just break, then we'd have big problems. Actually, too much metal flexing causes fatigue, and chassis platforms that ARE designed w/ flex in them are unusable after a certain period of time. For the loads most of us are putting on our cars, metal ductility and yield strength are not the issue here. Using thicker steel and structural foam does not change the characteristics of the material it is bonded to. A case could be made that adding reinforcements that are stronger than the material it is bonded to would cause the weaker material to yield, but again, we are not approaching high enough loads in most applications, and in others, if proper bracing is done at the attachment points, then it will not matter. If this were not true, cars w/ roll cages would rip themselves apart...there is a reason they don't weld a roll cage to the sheet metal on the outer/inner body of the car, for example.

For the last time, chassis stiffness is not SUSPENSION stiffness. They work together, but they're separate. You cannot have infinite weight w/ a mass/spring/damper system.

This is what our car's suspension looks like in simple terms.

M1 is the car frame/body/people.
S1 and D are the springs/struts.
M2 is the wheels and "unsprung" brakes that move w/ the wheels.
S2 is the tire.

If M1 flexes, it is UNDAMPED movement. S1/D are sized according to M1 and M2, to either minimize the movement of M1 (soft suspension along bumps) or minimize the time it takes to get back to its original position (harder suspension, resists movements).

I suggest reading this to help any of you learn how these systems work.
http://mathinsite.bmth.ac.uk/pdf/msdtheory.pdf" onclick="window.open(this.href);return false;


Now granted, this is horribly under simplified compared to real world application. But it gives us a basis for where to go.

Chassis flex is usually not noticed unless a car is running stiffer springs, or under higher loads. When the chassis does flex, it's spring rate is MUCH greater than that of the corresponding suspension. An infinitely stiff suspension has an infinitely large spring rate. Increasing a chassis' stiffness therefore increases its "spring rate", but only if it actually flexes a measurable amount does this come into play. Think of it this way, you have 2 car springs, one with 10#/in and the other of 500#/in. The 10#/in will show 50x the deflection that the 500#/in spring does. If the 10#/in spring is subjected to a 1 lb load, it will compress a measurable amount. The 500# spring in this instance will act more like a solid piece of metal, since the force is not great enough to move it a measurable amount. Chassis' are much like this. If a suspension is far softer than the chassis spring rate, it will work as it was designed. Obviously, as you increase the forces the suspension is subjected to, however, the larger spring rate can come into play. The previous example used a 10# spring, and a 500# spring, with minimal compression. What if the 10# spring were subjected to a 50# force? Well, this is clearly within the realm of a 500# spring now. As a spring compresses, it is exerting greater and greater resistance, and all attaching pieces have to be able to withstand this force without reacting, otherwise they become a spring themselves.

When a cars suspension changes from strut/spring to bumpstop, the effective springrate goes towards infinity. Well designed bump stops can help minimize this, but the spring rate still shoots up dramatically. For many people on soft suspension setups, this is the only time the chassis will flex since it is now the lower spring rate. For people in this category, increasing chassis stiffness will reduce some of the noises/vibrations associated with the flex, but will not gain a significant improvement in ride quality, as the suspension is doing its job. If the car is constantly riding on bump stops, than reducing chassis flex may make bumps "harsher" since the chassis is more compliant and will not bend, it would act as a go kart and the bumps would be transmitted directly to you. Here, having chassis flex would act as an immensely stiffer spring (albeit undamped). The only cars that are designed to ride on bump stops though are performance handling cars, so having a stiffer chassis would be more beneficial to handling, and thus negates the reason to have a flexible chassis. In no situation will have having a flexible chassis reduce noise since interior panels and everything else attached to the frame will move with the chassis, making noise.

Now, consider the car with a stiffer spring in place of constantly riding on the bumpstops. As suspension stiffness increases, the ability for the chassis to resist becoming a spring itself diminishes. If a chassis is twisting, the suspension components are no longer working as they should. So now instead of the tire and strut/spring being the only "springs" on the car, you have a third one which is the chassis. If a chassis is flexing and a suspension is not bottomed out, the suspension components that are in place to take said loan are not compressing as much as they should. You are effectively increasing the spring rate of the car this way, whereas if the chassis did NOT flex, then the spring rate would be less since the spring/strut are the only things deflecting. As we mentioned in previous posts, having higher spring rate means a harsher ride.

Consider it like this...if a chassis is stiff and does not flex, you can compare it to a block of steel. Mate a spring to a block of steel and compress it. Only the spring deflects and you have a spring rate equal to that of spring.

Now take the same spring and attach it to a stiffer spring. Now compress it. In the case of a soft spring like I mentioned before, the spring rate would be that of the soft spring. Since we know the chassis is flexing though, this cannot be the case...so the stiffer spring must also compress somewhat. If the bigger spring compresses, than the effective spring rate has changed compared to the single spring. If we can eliminate the flexing of the big spring, we can reduce the overall springrate, thereby improving ride quality.

What's more is that chassis' spring rates at flexure are not linear. As we can all agree, a stiffer chassis will promote better handling...but why? If you reduce the overall springrate as I discussed above, wouldn't you hurt handling? The answer is no. "Handling" is a cars ability to perform by what the driver expects. With a flexing chassis, the changing spring rate is unknown and undamped, so things do not transition well.

Also, please do not take me to mean dissipate, as in go away to nothing when I say absorb in suspension. A suspension SHOULD absorb all/most of the energy put into it, then project it back towards the ground.

And everything I have mentioned is either aiding in reducing the car's CG or has no effect on it. I do not want to build a roll cage.

This does not even begin to describe how much less noise is created, let alone vibrations.

Resources that support my claims:
http://www.valvoline.com/car-care/autom ... 20040301cs" onclick="window.open(this.href);return false;
http://www.stockcarracing.com/tipstrick ... index.html" onclick="window.open(this.href);return false;
http://www.b15sentra.net/forums/showthread.php?t=108496" onclick="window.open(this.href);return false;
http://www.fordfusion.net/specs/2006-fo ... hassis.php" onclick="window.open(this.href);return false;
http://books.google.com/books?id=_p-6jr ... rt&f=false" onclick="window.open(this.href);return false;

+Many more.

Please provide any information that supports your claims.

[300zxrb26dett]

Here is another good one

http://www.sonustc.com/la_zh/upload/Dow ... Basics.pdf

Its all good reading, but the part most interesting is toward the bottom where it says

Designing for NVH

Basic principles . That section spells it all out

[onlytrueromeo]

That is exactly what I'm talking about. Good find!

Maybe it will now put the issue to rest.

Quoted from one of the Flying Miata people:

"Being a unibody, a Miata doesn't have a separate frame like a truck. It does have pseudo-frame rails built into the unibody. But we're talking about the structural rigidity of the car. Separate the suspension (the moving bits) from the chassis (the thing the suspension is bolted to) and you'll get the idea. You definitely want some compliance in the suspension, that's what it does. But there's no reason to want compliance in your chassis.

This is important, separate the suspension from the chassis. You cannot have a chassis that is too stiff, regardless of where you use the car. If you're counting on the chassis to flex to make your ride comfortable, you've got a very poorly done suspension. I have a Miata that's seam-welded, fitted with a number of braces and has a full cage in it. The stiff chassis makes it much more comfortable to drive than a "normal" Miata.

In a strut suspension, side loads from cornering are transferred into the chassis through the lower control arm and the top of the strut. The top of the strut, of course, is at the top of an open box with fairly poor structural integrity. That's why a strut brace is so important on this kind of car.

The Miata, with a dual-wishbone suspension, feeds cornering loads in through the control arms only. The top of the shock sees the weight of the car and impacts from road imperfections, but not the sideways force under cornering that makes it so easy to deflect the entire front end. So the Miata doesn't need a shock tower brace quite as badly as a strut-based car, and that's where the "Miatas don't need an STB" statement comes from."

[Inodoro Pereyra]

Onlytrueromeo: don't get frustrated.

After reading these last posts you made, I seems evident to me that you somehow feel it is our duty to come here and "defend our position", when we don't agree with you.

So, let me make this clear: I don't have anything to defend. If I came here and offered my advise, is because I decided to help, not because I have any obligation with you or anybody else. I have, from the beginning, never found this thread particularly interesting, and, so far, the only thing I have learned from it is that you opened it not to ask for advise, but to look for other people's agreement. Well, I will not agree with you for one simple reason: because you're wrong. You don't need to be an engineer to know you're wrong. This is basic, high school physics.
Now, I have tried to explain myself with as many words as my limited English allows me to use, so, as far as I'm concerned, there's nothing else I can add.
You can, and obviously will, do to your car whatever you want to. You can fill it up with concrete, for all I care. You're ultimately the one that's gonna have to live with it. Either way, it doesn't change my life in the slightest.

[fowljesse]

Good discussion, everyone! I just wrote a bunch of junk to keep it going. I'm going to revisit this one when I get a chance to work on my car.

[Ryan]

What I get out of this, put in simple, non essay response...


If an infinitely stiff chassis were possible, it would be good.

Its impossible, but it would be good.

The optimum is a stiff/weight/cost balance.

In the end, I don't think I'd really notice if someone filled my frame overnight with foam.

[onlytrueromeo]

Onlytrueromeo: don't get frustrated.

After reading these last posts you made, I seems evident to me that you somehow feel it is our duty to come here and "defend our position", when we don't agree with you.

So, let me make this clear: I don't have anything to defend. If I came here and offered my advise, is because I decided to help, not because I have any obligation with you or anybody else. I have, from the beginning, never found this thread particularly interesting, and, so far, the only thing I have learned from it is that you opened it not to ask for advise, but to look for other people's agreement. Well, I will not agree with you for one simple reason: because you're wrong. You don't need to be an engineer to know you're wrong. This is basic, high school physics.
Now, I have tried to explain myself with as many words as my limited English allows me to use, so, as far as I'm concerned, there's nothing else I can add.
You can, and obviously will, do to your car whatever you want to. You can fill it up with concrete, for all I care. You're ultimately the one that's gonna have to live with it. Either way, it doesn't change my life in the slightest.

I'm not trying to bash you or piss you off, and I did open the post for discussion, but you don't offer anything but your word, and you say IM wrong. Please tell me how "high school physics" proves me wrong? I honestly want to know the answer to this topic. I have posted different links and the theory behind it, and talked to many fellow engineering students and graduates and they agree with me. I say that you are wrong AFAIK, because I cannot find any evidence nor do I understand the logic of where you are coming from. So far you are the only one arguing your position and you have offered no backing to this besides telling me I have no idea what I'm talking about. If you are right, there must be some reasoning/science behind it...I am not debating politics, I am trying to understand the science. I posted in detail why I believe I am right, and to the best of my ability I tried to spell everything out in laymans terms.

Filling a car w/ concrete is stupid. I'm not talking about reinforcing w/ brittle/heavy materials to increase the stiffness.

Also, if you can't explain in english, maybe you can use your native language and translate it through google? Sorry, I only speak English fluently, and know VERY minimal Italian.

[onlytrueromeo]

On a change of topic from chassis stiffening, but still of importance to this thread, I have been reading into increasing track width lately. I know the protege's use our suspension parts to increase their track, but we don't have any available cars to swap parts from. This means if we want to increase track in the same way, we would need to make custom parts! While it would be sweet to make custom control arms/trailing arms and lateral links, I don't have the capabilities or money to machine all this stuff.

By far the easiest way to increase track width and increase grip is to use wider rims/tires. When you deviate from stock setup though, you are affecting the suspension geometry, and the steering axis inclination (SAI) and the center of the tire's contact path. This is scrub radius, and on our cars we must have negative scrub, meaning the center of the contact patch is inside the SAI.

This picture is negative scrub, although it's not a perfect diagram, you can see how the line projected from the strut is OUTSIDE the center of the contact patch.



All things being equal, if you increase the width of the tire, the center will move closer to the SAI line, and will eventually become positive scrub as shown below.




Simply using camber plates will not change the SAI, we would need to modify the strut/hub mounts and the axis on which the steering pivots. I believe it can be done.

Normally I would say fender rolling is dumb, but after reading many posts/articles where people flared the fenders slight after, I think it could be done tastefully. The only examples of fender rolling I've seen on MX3's has resulted in a widening of the entire wheel well, which looks If you were to just roll the lip/flare it a little, you could accommodate a 1" wider wheel easily. Over the summer w/ my car weighted down w/ tons of stuff from school on my 15x6.5's w/ a +40 offset, I rubbed my tires on my rear fender on turns through the mountains. I had to go under the speed limit to prevent this. My car is not slammed, but neither am I stock - I believe I had B&G's on at the time, that or my H&Rs and if anything I had NEGATIVE camber which would make rubbing of my fenders even harder to do.


I will try to draw up diagrams later, but I have to get back to studying for a test tonight

[300zxrb26dett]

On a change of topic from chassis stiffening, but still of importance to this thread, I have been reading into increasing track width lately. I know the protege's use our suspension parts to increase their track, but we don't have any available cars to swap parts from.

Oh really?

Well let me be the first to tell you that yes you do have options, but only for the front.
FC RX-7 balljoints can be bolted on for increased track width and a little better roll center for lowered cars. The only problem is if your MX3 spindle will accept a larger balljoint stud, but if you dont, you have options for this also.

Read this entire thread, and the linked FEOA thread on the first page of this thread. All your questions should be answered, but if not ask me since it was me that discovered it.
http://www.clubprotege.com/forum/showthread.php?t=35403

[onlytrueromeo]

Hmm...I did not know this! I will have to do reading on it.

Thanks!

[fowljesse]

Thanks, 300zxrb26dett. That's cool!

[onlytrueromeo]

OK, I'm still learning so bear with me. By using the RX7 balljoint, we would be moving the hub out 1/2", and the control arm down a small amount since the shaft is a little longer.

In this picture, the balljoint shown would move the wheel hub (so tire/wheel included) towards the outside of the car.


Another pic:



This would increase our SAI. From what I have been reading (and I know very little about this so far) a higher SAI is not necessarily good. To counteract any "negative" effects of increasing the SAI, you would have to move the top of the strut out as much as the ball joint moved so that you do not change the angle. You would not gain any camber from this, but your track width would still increase the same amount. This can only be done w/ camber plates.

This is a really good article on the subject, and supports my belief that we should try to increase caster on our cars for the most benefit.
http://buildafastercar.com/tech/Dynamic ... nclination" onclick="window.open(this.href);return false;


So lets say we use the method of the RX7 ball joint to increase our track width (which I think is a good idea). From what I have read, it will benefit us most to use camber plates to adjust the top of the strut to reduce the SAI angle to stock, or even past stock, increasing camber. Camber can than be corrected by using camber bolts, which changes the included angle rather than changing the SAI. The car would benefit more from increasing the caster - moving the top of the strut towards the back of the car. This will increase our dynamic camber.

The problem with changing the SAI going back to my point of increasing the contact patch. If you decrease the SAI, you are then increasing the scrub radius. I don't think the mx3 will hurt from a slightly positive scrub, but I don't know this for sure. I would certainly want to keep the scrub radius as minimal as possible w/o being at zero.



Changing the SAI or the track width by moving the balljoint can also effect steering through the Ackerman Angle. I clearly do not have all the suspension parts in front of me so I can measure, but change the steering system could adversely affect the handling/grip capabilities of the car. You want the angle of the tires to match that of the curve they are on as best they can...


I will be doing more reading/research to add to this.
I am currently reading this to help improve my knowledge of roll centers vs the CG.
http://www.neohio-scca.org/comp_clinic/ ... cs2007.pdf" onclick="window.open(this.href);return false;


Please share your thoughts/experience!

[300zxrb26dett]

I lack the knowledge to really comment on this sort of thing to much, but figured this link will help some. I know what your trying to say though (I think)

http://www.hrsprings.com/technical/scrub_radius