Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

A Flexible Chassis is Slow 3

Status
Not open for further replies.

BUGGAR

Structural
Mar 14, 2014
1,732
This is relative to my investigation of the Ariel Atom chassis:


A Flexible Chassis is Slow.

In my quest to understand the importance of chassis stiffness of the Ariel Atom, I was talking with some off-road guys. We have a pretty big off-road community here in the desert and they’re a pretty open group. I was talking to them about chassis rigidity and they were talking about how much it slowed them down. What were they talking about? So here is a transcript of what I remember them saying, as best as I can remember and with some editing.

When you hit a bump, it’s like the bump is hitting your wheel with one of those little arrows that engineers use to show a force. It always points towards the center of the wheel because of the same laws that also keep people parallel to each other no matter where they are standing on the globe (sic).

Those force arrows pointing at the wheel can be broken down into two directions, one vertical and one horizontal. This is because engineers think in X and Y directions. In addition, those are the only directions that things can move when hitting a bump (called degrees of freedom by engineers – kind of like all the places where you can’t go if you’re under 21).

As noted, there will be an up arrow and a back arrow. The back arrow is related to and proportional to the up arrow. This proportionality changes as the wheel rolls over the bump but the back arrow is always going to be some proportion of the up arrow and it’s always going to be there until that bump is finished with the wheel. You can forget about the bump pushing back after cresting the bump because your car is flying off that bump without looking back for any help.

This back arrow forces back on the car’s suspension, and the car’s reaction is to slow down slightly. To reduce this slow-down arrow, we must reduce the up arrow, or more accurately, the vehicle’s reaction to the up arrow. This reaction to the up arrow comes from inelastic and elastic force/energy absorption: inelastic from the shock absorbers, elastic from the tires, springs and chassis flexibility.

For a given energy input from a given bump, the reaction force from the up arrow and its evil back arrow is less for an inelastic reaction than for an elastic reaction. Thus, the more energy absorbed inelastically by shock absorbers, the less the evil back arrow restraining force. If the chassis is flexible and does not permit the shock absorbers to absorb as much energy as they can, the resisting forces will be increased. That’s one reason to run multiple or progressive rate springs; they leave more work to be done by the damper than by the springs, and the slow-down arrows are smaller.


I think he’s right. I think. At least it was a great conversation!

Added by me: There are, however, minimum amounts of elastic energy that are required to restore the chassis and wheels to their neutral position. Also, experience shows that less than critical damping is “best”.


 
Replies continue below

Recommended for you

The difficulty in sensing via seat of the pants is understood, and I make no claims as to the resolution capabilities of my own butt-meter beyond picking up the coarsest differences. That said, I do have some datalogged information around a rather unofficial test loop where the only difference was in wheel widths and tire sizes (same tire make and model), plus (still) some memory of the differences in 'feel' out toward the maximum lateral g's experienced (0.7 - 0.8g or a bit more for the wider set).

On edit, this is with a car having a claimed 21,000 NM/deg torsional stiffness per the list linked to in another thread.


Norm
 
NoahLKatz said:
Regardless of the magnitude of the longitudinal horizontal forces, they are a significant contributor to impact harshness in suspensions w/o compliance in that direction.

I'm not sure what you're arguing, I don't think anyone is disagreeing with you.
 
I believe you're right.

I misinterpreted the below statement to say that because the horizontal force component is low, that it's not an important contributor top NVH.

"Point is, no matter how you try and rationalize it, the horizontal component of bumps forces is VERY small relative to the kinetic energy of the vehicle, and impulse applied to the chassis is controlled by the compliance of the suspension bits between the upright and the chassis, because those bits are going to be much, much less stiff than the chassis itself."

BTW, how do quote someone else's post?
 
Click the quote button, type the name of the poster you want to quote, copy and paste their text.

It's pretty clunky on this board compared to most others that use VBulletin or whatever and allow automatic quoting.
 
It's just as easy to just type the HTML for quoting straight out.

{quote name}<paste your text>{/quote}

Replace the above brace characters { and } with the respective brackets [ and ]. There is a space between "quote" and the name you want displayed as your source.


Norm
 
Thanks, Norm, but I'm lazy; that's too much extra typing and/or mousing around for me
 
"21,000 NM/deg torsional stiffness"

Highlight their text, ctrl c, then ctrl v it into a new post.

Having accomplished that, this is over 15,000 ft lb per degree?
 
Buggar said:
Having accomplished that, this is over 15,000 ft lb per degree?

Guess so. A little units conversion utility called 'Convert' came up with the same result. Probably 3 to 5 times what the 1960's originals were good for.


Norm
 
I use the same Convert. I guess if everybody's watch says the same, then that's the actual time.
 
Is there any data on this at all? I want to see how a controlled test on chassis rigidity's effect on speed when the vehicle travels over a given set of bumps.



"Formal education is a weapon, whose effect depends on who holds it in his hands and at whom it is aimed." ~ Joseph Stalin
 

What do you guys think of this? I don't think offroad truck guys actually know what stiff is to be honest. They might just know what "too soft" is. By the way, I have ridden the bike in that article at a competitive level in various conditions. If I had not ever done so, I would not be dissenting from the consensus in this thread. There are some things that are not readily apparent through the situations that you are currently capable of simulating in your head based on the experiences that you currently have/don't have. Do you have test labs where you work? Test labs exist for that very reason. If you disagree with that, it is because you miss understood what I just said, or you don't actually work in engineering.

"Formal education is a weapon, whose effect depends on who holds it in his hands and at whom it is aimed." ~ Joseph Stalin
 
Panther 140 said:
Is there any data on this at all? I want to see how a controlled test on chassis rigidity's effect on speed when the vehicle travels over a given set of bumps.

Maybe someone has tried... In my opinion it would be very very difficult to conduct a test of this interaction where the actual effect of chassis stiffness is discernible from noise in the experiment.

Panther140 said:
What do you guys think of this?

I think that motorcycle chassis and full scale off road vehicle chassis have very, very little in common. I don't debate that for the particular Honda motorcycle described in the article, what the author says is true (the gist of which is, this bike sucked because the chassis was too stiff) but that has zero bearing on a trophy truck or sand rail or whatever.

Panther140 said:
I don't think offroad truck guys actually know what stiff is to be honest.

This is (with, perhaps, the exception of the top .01% of the sport) very likely to be true.

Panther140 said:
They might just know what "too soft" is.

I doubt this strongly. To know what 'too soft' is with regard to an off road chassis, you would have to build multiple off road chassis of known stiffness, with the stiffness of each varied by a statistically significant degree, outfit them with the same components, and then test extensively. I doubt VERY strongly than any off road racing team has ever done this, with the explicit goal of trying to determine the exact effect that chassis stiffness has on suspension tuning. In my experience, they (attempt to) build things so that they don't break, and then tune the suspension to suit their attempt at a bomb proof chassis. There's very little engineering-from-first-principles going on. That's not a knock against those teams- it's a tough environment to design in.
 
There are rules of thumb for "too soft" or not. Basically the wheel rate attributable to chassis flex should be at least (10x ?) the wheel rate attributable to springing components (springs and ARBs).

je suis charlie
 
As Milliken nearly said, if you are designing thumbs, use rules of thumb, if you are doing vehicle dynamics, perhaps not.

The problem with that one in particular is that correcting the springs for a soft chassis is easy enough, schoolboy physics. The hard part is figuring out how to modify the shocks.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
gruntguru said:
There are rules of thumb for "too soft" or not. Basically the wheel rate attributable to chassis flex should be at least (10x ?) the wheel rate attributable to springing components (springs and ARBs).

Right.. but how many guys building a sand rail in their garage, or how many custom shops building a sand rail for a rich guy to park in his garage, know their design's true wheel rate due to chassis flex?

My guess would be close to 0.
 
Greg. I think that particular rule-of-thumb is meant to be a standard for building a back yard chassis rather than a tuning tool for compensating a sub-standard design.

je suis charlie
 
ps: the 10 percent generally applies to torsional rigidity of the chassis vs. the roll stiffness of the spring/anti roll bar stiffness? Correct me if I'm wrong. Is it valid? See my posts on the Ariel Atom. But that's another subject.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor