Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

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

Ariel Atom 1

Status
Not open for further replies.
BUGGAR

BUGGAR

Structural
Mar 14, 2014
1,732
A member on a car forum sent me some dimensions for the chassis for an Ariel Atom for me to run a structural check on. Has anyone else done this? This chassis is torsionally very flexible for a performance car (under 1000 ft lb per degree). Is a torsionally soft chassis now considered better?
 
Replies continue below

Recommended for you

Sort by date Sort by votes
Excellent write up. Thank you.

If your spring rates at the wheels are anywhere near 50 pounds per inch, then the chassis I'm working on meets the 10 times "rule" (chassis should be stiffer than 10 time the roll stiffness of the suspension), and all is good.
 
Upvote 0 Downvote
Adjusting dampers to tune the F-R distribution of transient roll resistance relies on MOI(x) reacting the transient roll moment at each end.

With a torsionally stiff chassis, both F and R roll moments are acting on the full MOI and both ends will be quite sensitive to damper settings.

With a very floppy chassis, F and R will each react against a different value for MOI since the front and rear of the chassis are decoupled to some extent. (imagine a chassis with a longitudinal hinge joint in the middle of the chassis) So a car like the Atom (mid/rear, transverse engine) will have a much larger MOI connected to the rear axle than the front. The consequence is much higher sensitivity to rear damper settings than front.

je suis charlie
 
Upvote 0 Downvote
buggar said:
If your spring rates at the wheels are anywhere near 50 pounds per inch, then the chassis I'm working on meets the 10 times "rule" (chassis should be stiffer than 10 time the roll stiffness of the suspension), and all is good.
Click to expand...

Suspension roll stiffness (a moment stiffness, just like chassis torsional stiffness) needs more information than just the spring rate (a translational stiffness, reflected to the tires). Big hint: things like consistent units and track width probably have something to do with it.


Norm
 
Upvote 0 Downvote
As a matter of interest what are the dimensions of the four main tubes? And the wheelbase? I assume they're aluminum.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Upvote 0 Downvote
NormPeterson said:
things like consistent units and track width probably have something to do with it.
Click to expand...

The other thing about an Atom is that it's a pushrod car- so if the design relies on the main springs only to provide roll resistance, without the aid of a sway bar that would act in a linear fashion relative to differential suspension displacement, than the roll couple is also affected to some degree by where the car is in its suspension travel- because the spring behavior, when applied through a bell crank and pushrod, is less linear than, for example, a MacPherson strut car without sway bars might be.

This is evident in tuning an Atom- you can directly and noticeably impact the weight transfer rates (in roll) with small ride height changes. This is also useful as a tuning tool.

Also, BUGGAR: it looks like the car currently has ~40 N/mm springs front and 45 N/mm rear.

I fooled around just now with MS paint and a photo of the rear of an Atom, estimating where the rear roll center is (this thread is making me curious) and it appears to be approximately half way between the floor pan and the ground. The real question is where the CG is located. I suspect it's relatively low, but it's hard to estimate.

The low rear roll center doesn't surprise me too much- the car is so over-powered relative to its weight, I suspect it would be very twitchy with a high rear roll center.
 
Upvote 0 Downvote
Wheelbase: 103", track: 60" (specific vehicle I am working on).
Main tubes: 2" x 0.065 wall, d/t = 30.
Tube material: 4130(or 4140?) steel. One manufacturer says DOM, another ERW.
My scope right now is elastic rigidity analysis only; analysis for overstress won't occur if this chassis can't prove itself more rigid. The so called 10 times rule is just a reality check. I can present all my Risa files but independent verification would be better.

I'm not getting paid for this; I thought these cars "looked cool" and wanted to try engineering one. Probably a very bad life choice.

I still think these cars are structural weenies but if this is the way to go fast, I would like to know this.

 
Upvote 0 Downvote
That was another question: Why use helper springs (I presume for a progressive rate) when you can set rate progression with bellcrank angles? I noted you found it desireable to eliminate the rising rate springs.
 
Upvote 0 Downvote
BUGGAR said:
That was another question: Why use helper springs (I presume for a progressive rate) when you can set rate progression with bellcrank angles? I noted you found it desireable to eliminate the rising rate springs.
Click to expand...

I don't know for sure but I believe the helper spring setup is only present on the more streetable versions of the car, to provide at least some compliance on the road. I'm pretty sure that once you 'upgrade' to a more track-oriented option set, the helper springs go away. I also suspect that on-road manners are the reason why initially the car was underdamped. As I said earlier, right off of the showroom floor, it was a mildly scary car mid-corner on a bumpy course because it was already so close to being on the bump stops steady state. A simple increase in front bound damping (within the range of the shocks- nothing has been re-valved on this particular car) was enough to kill this behavior.

BUGGAR said:
I still think these cars are structural weenies but if this is the way to go fast, I would like to know this.
Click to expand...

It seems that 'structural weenie' is an apt description.

I think this is A way to go fast, maybe not THE way. This car is so stinking light, and is vastly over-tired and over-powered, that I think it gets away with a lot with regard to the chassis being less than optimal. If the goal was to turn an Atom into a fast autocross car that had to compete at a similar power-to-weight ratio as some other cars, I suspect it would be very hard to make it bleeding-edge fast.
 
Upvote 0 Downvote
buggar - I understand the sanity check nature of "10 times". But you need to start with the right number before you multiply it by 10 or else it's a case of GIGO.


The formula is almost as easy to derive as it was to find it online. Personally, I hate having to swap back and forth between feet and inches so I keep it all in terms of inches and use 0.008727 as a numerator coefficient instead of the 1375 you'll find in the link.


Norm

(edited for boldface formatting)
 
Upvote 0 Downvote
"I think this is A way to go fast, maybe not THE way. This car is so stinking light, and is vastly over-tired and over-powered, that I think it gets away with a lot with regard to the chassis being less than optimal. If the goal was to turn an Atom into a fast autocross car that had to compete at a similar power-to-weight ratio as some other cars, I suspect it would be very hard to make it bleeding-edge fast."

This, too, is my conclusion after beating up on this subject. Maybe I didn't learn much but now, maybe I know in which races I should compete with these things.
 
Upvote 0 Downvote
BUGGAR,

Wouldn't the total mass of the car affect the way the car flexes? Is any street car anywhere near the mass of an Arial Atom?

--
JHG
 
Upvote 0 Downvote
Caterham, various locosts, original elan, original mini, mescersmitt 3 wheeler, lots of austin 7 derivatives.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Upvote 0 Downvote
I'm targeting torsional rigidity, independent of mass. The mass plays in indirectly through the spring stiffness requirements and the resulting suspension roll stiffness, which is where the 10 times factor comes in.

Norm, I am using the spring rate at the wheels so I use the 60" track width between spring elements (wheels), which will be lower than the actual spring rate. Let me know if I pulled a Nancy (Sinatra - sayin' somethin' stupid).
 
Upvote 0 Downvote
GregLocock,

How about I modify a Dodge Challenger (4000lb) so that it has the same performance as an Ariel Atom. Obviously, I need way more torque and power. The suspension is going to see way higher loads side to side and fore and aft. The need for suspension rigidity should be very much greater.

--
JHG
 
Upvote 0 Downvote
drawoh said:
Wouldn't the total mass of the car affect the way the car flexes? Is any street car anywhere near the mass of an Arial Atom?
Click to expand...

Structural frequencies would be affected (and not expected to be well damped). I don't think the Mark I Sprites were much more massive, either.


Norm
 
Upvote 0 Downvote
Drawoh, sorry I'm missing your point. You asked a question. I answered it, off the cuff. Here's another stab-

If you had two cars otherwise identical, and merely turned the engine power up on one, then I doubt you'd worry about stiffening the body as a first step. Far more likely you'd fit bigger tires and retune the suspension to be sportier, in which case a stiffer body would help, due to the whole springs in series idea.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Upvote 0 Downvote
BUGGAR said:
I am using the spring rate at the wheels so I use the 60" track width between spring elements (wheels), which will be lower than the actual spring rate
Click to expand...

50 lb/in on a 60" track gives something like 130-ish ft-lb/deg roll stiffness, against less than 7.6 times that in chassis torsional stiffness.

Looked at from a different angle, 50 lb/is wheel rate is pretty soft for the intended purpose of an Atom . . . suspension ride frequencies in the vicinity of 1.1 Hz and an implied roll rate that could be as much as 3.5°/g both seem softer than what I've seen from an Atom on track.


On edit, most laps, that's about a 1.1g corner.


Norm
 
Upvote 0 Downvote
Is that 3.5 corrected for gravity and tire stiffness? I repeatedly run into problems with the latter, as it is such a large proportion of the total roll compliance number. I'm not really a fan of the total roll gain number for a circuit car especially, suspension roll is far more useful as it feeds directly into your roll steer and understeer budgets.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Upvote 0 Downvote
That was just a very quick first cut, taken as an estimated roll moment divided by total suspension roll resistance only; I didn't even go into my datalogs to line up that particular corner on that particular lap with speed or g's. I'm guessing that tire compliance effects would add maybe half a degree per g to that, as an Atom likely has relatively stiff tires compared to its weight and track width.

So for the Atom in that picture, perhaps 4°/g or somewhere between 4° and 4.5° total is what I'd expect from 50 lb/in wheel rates and no sta-bars - or more than it looks like the Atom in my picture is showing. That 1.1g comes from actual datalogs taken in my own car, corrected downward for about 3.1°/g roll, which includes about 0.7°/g in tire effects.


Norm
 
Upvote 0 Downvote
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top Bottom