LocostUSA.com

Ah, but if you make the upper tubes with the long side horizontal, it cuts into the very limited seating area, which kind of rules it out from a practical standpoint. I agree that it makes sense to help close off the open top box structurally. I intend to use a low front cage, which probably will stiffen up the whole cockpit area enough not to need further reinforcement, so I should concentrate my efforts on stiffening the engine bay.

For all of you who are using grape (or other programs, i just dont know how accurate the stress numbers are for grape) what is the maximum allowable stress that you model for? With a 2000lbs torque on the frame, most of this chassis shows 13,000 psi of stress. Does any one else look at stress, and what values do you go for with what loading?

click to see full size
currently 236lbs weight and 10,085lbs/degree.

It's good to see another person taking a stab at this. I'm not an ME so take my comments/questions with a grain of salt.

Since the members of the chassis are colored uniformly it looks like you have used each tube as an element? I note the other simulations appeared to model the tubes with some or many elements down the length of the tube.
I think the actual stiffness result would depend a lot on that. Your stiffness number is just considering the elongation and compression of the tubes, no bending is taken into account.

Does Grape do the meshing for you or is that done by hand?

Obviously, you need to keep the calculated stress under the material yield point. The hard part is making correct choices about assumptions for loads and the suitability of your model...

Please keep it up, we all like pictures!

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.

each node is hand entered, mostly off of a sketchup drawing I am working of of. each element is straight, from weld point to weld point, (nodes), and should not need to be modeled in multipull pieces. With 2,000lbs loaded on the right front corner, and left front restrained, with rear points restrained in y only, the highest axial load (load down any indivdual tube) is 4673lbs. stress of 13,000psi, vs 31,900 psi yield for hot rolled 1025 (mild steel) (saftey factor 2.45 to yield) or 53,700 psi cold drawn 1025(saftey factor 4.13 to yield) (from http://www.matweb.com). but 2000 lbs load on a single corner is hopefully unrealistic for a 1300-1800 lbs car. hear is axial loading pic



I am looking into the calculations for small column buckling, but as I never got to ME classes, I am having a hard time with figuring out the proper formulas/ how to work them (euler, jb johnson).

any one with experiance in this area, a tutorial would be exelent!

In theory, your finite elements can treat each tubed polygon as a single element. But I don't think it works that way in practice for a typical real world space frame. You can see this in the other examples in the thread, the color of the tube changes along it's length. Indicating that the tube is being stressed by bending. From the pictures, it seems to be a very significant effect. I haven't seen anybodies data files so I don't know how the meshing was accomplished.

It's good to be conservative about assumed loads, even on a track only car. There was a discussion on the UK Locost site about what factors to use, sorry I don''t have the thread at hand. But the engineers there spoke of people who had done actual testing. You could calculate what happens when you cross a 1" step at 100 FPS. It's a good thing tires have sidewalls...

So, I think the 2000 lbs. is a little light when you think of the effect of the shcok absorbers damping in response to road bumps... And the 10,000 lbs. of chassis stiffness is probably much less when modeled closer to the real world. Don't be discouraged by this - this is where the real learning comes in. Improve your assumptions and models and that's when tremendously valuable lessons are gained, the real payday so to speak.

Please post your data files at some point, that will help others with their learning curves.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.

the impact of a tire on a pothole rising edge, 2x4 or similar is something that i was looking at calculating. a 1" step hits the tire 4.85 inches before centerline, (12.3inch radius), with a vertical loading of around 360 lbs, but i dont know how to calc the z (front to rear) componet, or realisticly the y componet either. should have some thing to do with the curve of the tire, max acceleration in vertical direction would hapen at point of impact, then decrese as tire rolls over, im shure a calculus limit problem is needed here, but I dont know how to solve it.

grape does calc deflection of each node due to bending of elements, but it is such a horible picture with this model, green dotted lines on top of an already jumbled mess of green lines, you cant follow anything, other than to look at the displacements, node by node.

I recalced the model for 3,600lbs (10g at 360lbs corner weight) at rf corner (model has front suspension, with solid link for shock/spring.) max axial load went to 8,412lbs in one of the front dashboard elements, with max axial stress of 26,804 lbs. displacement of spindle where load is applied displaced 2.4009 inches y direction. 68 inches from load point to restrained front point. 2.0221 degrees rotation, 10,088 lbs per degree.

i think that in torsion bump case, frame stiffness will lift whole frame, including rear in responce to 1 wheel bump, and not be restrained at the rear in vertical direction as in model, lowering observed loads and stresses in the chassis. on a real life 1800lbs vehicle, 3600lbs vertical on right front will probly just flip the car over, and not bend the frame very much!!

So the heel must rise 1" in .004 seconds to clear this obstacle. No calculus. But you've reached the limit of my math, and we probably don't know a value for the shock stiffness. If we had a dyno chart for a shock, we might be able to get a guess value.

When cornering, the car should be able to transfer a lot of weight to the outside wheel. I would just approximate by using %50 car weight for the load on a corner.

I am sorry to pull a number out of the air, but I recall something like 3 or 5 g's being talked about in the Locost UK forum for calculating suspension loads on bumps.




In the dynamic sense, that movement is caused by the max. force you are calculating. The inertia of the frame supplies that fixing - at least with regards to instantaneous forces.

I'm just thinking out load with you, not really an expert - enjoying what your doing though.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.

I was thinking the same thing, under hard cornering + braking, that outside front will be carrying most of the car weight. And I have also heard of using several additional (like 3 to 5) g's of load to account for bumps.

But I don't think having the frame yield is really a concern, so I would just compare the torsional rigidity (in pounds per degree of twist) to other cars with similar weight and purposes.

OK, here's the deal...I'm not an engineer. My degree is in forrest management and I'm a retired Firefighter.

I'm building a McSorley 442 with a 3.4 V-6. I was wondering if I needed to do anything to beef the chassis for the additional torque the V6 produces. I started reading this thread last night and somewhere on the second page I began to drool and fell asleep. My wife woke me at 11 pm and sent me to bed.

Is it alright if I just ask...is there someone on here who has built the 442 with said drive train and did the chassis hold up to moderate floggings?

_________________
RacerDan
+442 Chevy 2.8
Never be afraid to try something new. Remember, amateurs built the
Ark, professionals built the Titanic.

as far as additonal torqe from the engine is concerned, i doubt it, but i dont know about additional engine/trans weight.


I came up with 6,785,277 lbf of force in the first 1/10,000 seconds of the impact in the vertical direction, but that only lasts for 1/10,000 seconds, and it only involves a vertical movement of .07405 inches if the tire was incompressable. For some reason, I think that that load would never make it to the frame, due to such a small amount of movement in the shock along with short duration. I still dont know how to figure out the rearward loading of the spindle, that concerns me more as there is no dampening in that direction with rodends, and there is the potential for the spindle/upright/control arms to see the compleate weight of the vehicle on one wheel (think hitting a large curb with one front corner while sliding--- would perfer control arm colapse to control arm to chassis mount failure)

From what I can find the V6 is only 40# heavier than the Miata and 65-70# lighter than the 2.3 Lima/Pinto I4. That tells me I'm good to go.

_________________
RacerDan
+442 Chevy 2.8
Never be afraid to try something new. Remember, amateurs built the
Ark, professionals built the Titanic.

Slngsht here and (USA7s usually) has a 1" square tube chassis with a LS1 and it seems to stand up to flogging just fine.

_________________
He is a wise man who does not grieve for the things which he has not, but rejoices for those which he has.

I think that 5 was vertical, 4 was longitudinal and 3 was lateral - but the autosport thread should clarify that.

You can find the most recent thread on the subject on the UK locost site here:

http://locostbuilders.co.uk/viewthread.php?tid=80810&page=1

It dries up when we tried to establish how to introduce these load cases into the model. Since then I've managed to get a copy Costing & Phipps and can see that the technique that he uses is to apply the acceleration to the whole chassis, and then establish the resulting acceleration of the complete car (assuming no chassis deflection). He then replaces the various masses with the forces resulting from accelerating them. These forces are applied to the chassis through the appropriate mounting points. Hope I'm not being unintelligible here!

Cheers,
James

_________________
Before you judge a guy walk a mile in his shoes. Then when you judge him, you've got a mile head start and you've got his shoes on:)

WHAT??????

_________________
RacerDan
+442 Chevy 2.8
Never be afraid to try something new. Remember, amateurs built the
Ark, professionals built the Titanic.

Sorry Dan, which bit! WHAT??????

_________________
Before you judge a guy walk a mile in his shoes. Then when you judge him, you've got a mile head start and you've got his shoes on:)