LocostUSA.com

When designing chassis components and analysing it using FEA what are the main points to conceder?
I assume that the main part of the chassis is designed to give maximum torsion stiffness between front and rear suspension pickups as well as the stiffness along the chassis front to back.
How would the front uprights be analysed?
In a straight line there would be a vertical load equal to the corner weight+ bit of horizontal friction?
During braking there would be a horizontal force equal to approx half the car weight times the braking G force on each upright plus the vertical load of the car?
During cornering the outside wheel would produce an axial load toward the centre of the car through the hub/ bearing and a vertical load of the corner weight?
How close am I in my assumptions? Can someone p[lease point me in the right direction?
As the car I am working on will use F3 size slicks and weigh about 550-600kg I am assuming a conservative braking Force would be 2G and cornering would be 3 G.

You have the right idea.

The chassis torsional test is really just to make sure the torsional spring connecting the front and rear suspensions is stiff enough. If it is not then it will be the dominating spring rate and no matter what spring or bar combination you use the car cannot be tuned. Its pretty simple since you input a force at a point to create a torque and you determine how many degrees the chassis twists. You can use 500 lbs, 1,000 lbs, 2,000 lbs, etc and the chassis spring rate will be the same.

Figuring out the suspension forces is much more complicated. You need to know the lateral and longitudinal forces at the contact patch. These can be found by determining the static weight then adding or subtracting the weight transferred by cornering, braking, or accelerating (equations are available in pretty much any suspension book or on the internet) - multiply this weight by a rough shoot from the hip guess* of the tire's coefficient of friction (about 1 for non-aero and street tires, 1.5 for race tires and non-aero, and 2+ for race tires and aero). You also need to account for bump loads of about 5G to make sure a big bump will not cause a failure. The vertical weight will need to be multiplied by this vertical acceleration. I'll typically look at numbers for normal driving (-1G vertical) and a maximum bump G. As long as the normal driving numbers don't exceed 1/2 of the material's yield stress, fatigue probably will not be a concern (I will typically add an addition factor of safety here to make sure there is enough margin). Then I make sure the part will not break under 5G by comparing the stress to the ultimate stress values.

From there you figure out your lateral, longitudinal, and vertical ball joint forces at the upright by summing moments and forces to be zero. These ball joint forces must then be transferred through the arms - using the same static equations that all forces must be zero, you can insert the ball joint forces and back out the forces at the inboard attachment points. Lastly these inboard attach point forces will be the same for the arm and the frame (since the sum of all forces must be zero). Once you know the X, Y, and Z forces at all points, you can run some models to find what you are looking for.

The math isn't extremely hard - it is stuff typically seen by ~2nd year engineering students. If you are not an engineer but are good with math, finding a second hand Statics 1 or 2 book on Amazon it will help greatly. You just have to be very careful calculating what you are doing since making something positive instead of negative will ruin your results.

*bribing someone in your favor tire manufacturer's engineering department would also work as would paying a 3rd party to test the tire - neither are really feasible for one person building in a garage

test, since 12+ hr after my last (and first) post, it hasn't appeared.

Moderators: (I didn't see how to contact either of you...) I presume this is because new members undergo a probation period; how long is it until posts appear immediately?

Thanks

Noah

You need 2 then you're free to post.

Ah, good, thanks.

I've also been thinking about how to go about upright design using FEA.

I'd just as soon buy them, but there just doesn't seem to be anything not overstrong (i.e. heavy) for an ~800 lb car (3 wheeler, 2F/1R).

Even ones appropriately sized for the loads are overweight because structural inefficiency is sacrificed for manufacturability (same goes for CNC'd though not as bad); I believe the lightest upright would be be a monocoque structure which could be constructed by two intersecting tubes, one round for the bearing shaft and one rectangular for the body.

The latter could be further optimized by folding/rolling sheet material and welding at the seam.

But back to the topic.

Andrew, why would you bother with error-prone hand calcs when you could do it all with FEA?

Suspension and steering links could be modeled with beam elements (with appropriate end releases), which would be connected to the uprights and frame.

On one side the presumably fine mesh upright could be replaced with a rigid element.

Assign appropriate mass to frame, which could be a few solid or plate elements.

Connect the wheel axis to (the) ground with a beam element to react loads, which are applied as the usual bump/cornering/braking G-load cases.

Then correctly proportioned simultaneous loads would be applied to the upright at all of its connection points.

What do you think?

Might you have a FEM (NASTRAN .dat format) of suspension you could send me to play with? If so, my email is in my profile.

I also want to look at replacing rod ends/needle bearings with flexure pivots, which I have quite a lot of experience designing, though not for cars.

Thanks

Noah

Have you checked Pegasus racing for formula car units? Not so cheap though. There are also nice pieces available form 600racing and another outfit that does small circle track cars. You can also look at pictures of parts for dirt circle track cars, there is a variety of choices there for ideas. They are as you describe with t he addition of sheet metal gussets for the steering arm. Those types of places also sell the tapered steel inserts for ball joints and the axle stub to fabricate your own upright.

There are also some pictures on this site, you just have to read it all to find them!
Triumph Spitfire uprights are around 5 lbs., how light do you want to get and is this the place to find the last pound or two? I understand the Pinto and Miata units are too big...

I came across Pegasus in my googling; pricey and inefficient flat topology.

Thanks for the tip on 600racing; I emailed them for weights of these (reasonably priced) parts http://store.uslegendcars.com/site/depa ... 34A0874BC2

I have looked at various dirt track sites, but those uprights are mostly designed with kingpin bushings for live axles, and seem to be low very low height.

Another pet peeve I have is the use of solid axles, whose small diameter necessitates a heavy section to take the bending loads.

Also there's no reason to have any physical structure on the axle centerline; it just creates a longer load path from tire to wheel center and back out through the upright to the wishbones.

So why not a much larger, say 2" dia, hollow axle using thin section bearings?

Also why does no one seem to use the thickness of the upright to space the bearings farther apart?

Fair question about the weight.

I think a fabbed upright could be ~2 lb. Would 3 lb less out of ~40 lb unsprung weight be perceptible? Beats me.

What we really need is lighter tires with load capacities that aren't 2 - 3X more than necessary.

Then pick smaller, skinnier tires, but you'd notice the lost grip as the tires' limit is closer and passed more easily. Is the intent minimum weight/maximum mileage, or having something brilliant to drive? Reducing sprung mass is nice, but there are much better, less compromising ways to keep it lightweight, or at least as lightweight as it realistically needs to be without negatively impacting the roadholding or safety. Predictable, capable handling isn't easy to achieve with a trike configuration to start with, I'd rather lose the 10lbs myself than take it from the tires.

There aren't really any decent 13" or 14" tires that I know of.

In 15" the Toyo Proxes T1R 195/45R15 at 15.7 lb is the lightest I can find.

So the 800 lb 3-wheeler will have tires that weigh half as much as the ones on my car, which weighs 4X as much, and has more cornering power than I want to use (they are Bridgestone RE-11's though), but I suppose I'm out of the norm here as I'm not interested in racing.

There are decent 13" and 14" tires out there, but most of them tend to be either a budget mild performance street tire or R compound track or autocross tires. For the budget friendly skinny 13" tire designed to fit 5" to 6" wide rims check out the MG, Triumph and vintage Lotus forums to see what tires are available. For the more track and autocross friendly tires look to either the Toyo RA-1, if you anticipate light rain use, or Toyo R-888 if you only plan to run in dry weather. If you were interested, you could run either Toyo as a road tire, but they would kick up lots of debris due to their sticky nature. Also expect them to become heat cycled out after a year or two of use; probably well before the tread is worm out. Or look to the Hoosier A6/R6 line for hardcore track & autocross use.

My Europa will be driving on Toyo tires when it gets revived and put back on the road.

Check out this place for oval track supplies. I think they are the real deal, it's a really big catalog. For instance in the chapter linked below they sell fabricated sheet metal front uprights and also ball joints. Not just ball joints, but about 75 part numbers for precision rebuildable ball joints. And then they also sell gauges to check your ball joints.

The spindles are a few pages down. They may be too heavy for our use, though I can't tell the dirt oval track people seem very weight conscious. Lot's of titanium and magnesium....

http://www.behlingracing.com/catalog/Catalog08pgs83-92.pdf

Meant to say above that the spindles in the catalog show a pretty easy construction...

I have searched on 13" tires in the British car forums and found a lot of posts, only to discover that most are no longer produced.

I think there were one or two left, but I'm not going to bet that they're still available by the time I actually need them.

I looked at the Toyo's but the lightest R888 is 17 lb and the lightest RA-1 is 20 lb.

Marcus, thanks for the link; very interesting catalog but all heavily built stuff for American-based cars.

Unfortunately those tend to be the limitations we are faced with. You could probably run motorcycle tires if weight is your main concern, but I'm not sure how well the contact patch would be shaped for use in a non-tilting application. Also if I'm not mistaken, many motorcycle tires are very soft when compared to car tires so would probably need to be replaced more frequently.

The slicks and wheels used on formula cars and sports racers usually tend to be very light, and would probably fit your load range, but would be completely unsuitable for road use and would wear out at a truly amazing rate.

Have you looked into what wheels and tires are available for the original Mini?

Unfortunately everything about a road car is ultimately a compromise, especially if you are working within a budget. It requires needing to satisfy a large number of requirements that sometimes conflict and doing so in a reasonably economic manner. Starting with parts manufactured by others and readily available used is usually a big step towards satisfying the economic limitation. Custom uprights and ultralight suspension components are possible, but usually either expensive to CNC (if you don't have access to the equipment) or hard to ensure proper safety margins are being met.

The problem with suspension failures is that, by their general nature, they tend to happen at the worst possible time. Usually the suspension is loaded to he breaking point when something fails, and then the rest of the structure suddenly needs to deal with the outcome of the resulting failures; frequently leading to impacts (terrain and possibly other vehicles and structures) and loss of control. Personally, I'd prefer a bit of extra safety margin; I've seen the end results of suspension failures at speed. It usually ends badly.

Aren't the original Mini's 10"?

I think that will be going to far in the small direction; pretty hard to fit brake discs into etc.

The Spitfire upright is a reasonable place to start, though it's hard not to fall into my unreasonable ways and want to design/build everything much closer to the ideal.

Re suspension part breakage, those are good points.

I'd address that by conservative design (including fatigue analysis), using reasonably ductile materials that don't suffer brittle failure, and periodic inspection.