Peugeot 205 Saloon Libre - RWD Space-Frame Silhouette Racer
Discussion
I'm not sure it's necessary, to be honest.. I certainly don't have the budget for it and I'd be surprised if the likes of Radical / Caterham did much more than a visual inspection of welds on top of prototype durability testing.
Since the whole car is a prototype, the durability testing on all components will be when I come to drive it. That's when I also verify my assumptions on loading (i.e. g inputs) etc, and that's about as good as you can get at this sort of level! I could test every welded part, but it would set me back years.. I think it's better just to have a sensible safety factor at the design stage.
As you said, I think pipelines / aero / civil engineering projects and the like are subject to extensive tests because of the consequences of a part failing.
Since the whole car is a prototype, the durability testing on all components will be when I come to drive it. That's when I also verify my assumptions on loading (i.e. g inputs) etc, and that's about as good as you can get at this sort of level! I could test every welded part, but it would set me back years.. I think it's better just to have a sensible safety factor at the design stage.
As you said, I think pipelines / aero / civil engineering projects and the like are subject to extensive tests because of the consequences of a part failing.
CamMoreRon said:
I'm not sure it's necessary, to be honest.. I certainly don't have the budget for it and I'd be surprised if the likes of Radical / Caterham did much more than a visual inspection of welds on top of prototype durability testing.
Since the whole car is a prototype, the durability testing on all components will be when I come to drive it. That's when I also verify my assumptions on loading (i.e. g inputs) etc, and that's about as good as you can get at this sort of level! I could test every welded part, but it would set me back years.. I think it's better just to have a sensible safety factor at the design stage.
As you said, I think pipelines / aero / civil engineering projects and the like are subject to extensive tests because of the consequences of a part failing.
I'd suggest just carrying out a basic 2x overload test on the part, should be pretty easy with a simple steel frame rig and a trolley jack etc!Since the whole car is a prototype, the durability testing on all components will be when I come to drive it. That's when I also verify my assumptions on loading (i.e. g inputs) etc, and that's about as good as you can get at this sort of level! I could test every welded part, but it would set me back years.. I think it's better just to have a sensible safety factor at the design stage.
As you said, I think pipelines / aero / civil engineering projects and the like are subject to extensive tests because of the consequences of a part failing.
Assume worst case, ie complete car mass at max lateral g, add 50% for impact loading and then double it!
TBH, i think you just need to regularly inspect those parts (and in fact the complete spaceframe) during early development running, and being fabricated steel sheet construction hopefully any failure would be quite graceful rather than catastrophic!
I don't want this to come across the wrong way and make myself sound like a pompous t
t, as I do appreciate questions, opinions, and feedback; but suspension design is my day job at an OEM and we don't even test to those extremes on commercial vans!
I have conservative enough predictions of 2g in braking / cornering and 4g in bump, and safety factors on top of that. There's no extreme customer use to be allowed for, so if I brake in to a massive pot-hole or smash a square kerb then I will expect something to get bent or broken - that's just life!
The FEA simulations I've done are pretty comprehensive, as is how I've calculated all the input forces & constraints, so I've got good confidence in them. I won't be going and immediately hammering up the hills.. as you rightly said I'll be sure to shake it down gradually and ramp up to "maximum death" after each check.
t, as I do appreciate questions, opinions, and feedback; but suspension design is my day job at an OEM and we don't even test to those extremes on commercial vans!I have conservative enough predictions of 2g in braking / cornering and 4g in bump, and safety factors on top of that. There's no extreme customer use to be allowed for, so if I brake in to a massive pot-hole or smash a square kerb then I will expect something to get bent or broken - that's just life!
The FEA simulations I've done are pretty comprehensive, as is how I've calculated all the input forces & constraints, so I've got good confidence in them. I won't be going and immediately hammering up the hills.. as you rightly said I'll be sure to shake it down gradually and ramp up to "maximum death" after each check.
Interesting. I'm about to undertake a similar project with a 205 body on a shortened Impreza floorpan and restrengthening the tub post welding is a concern.
It'll have a cage which will help overall but the floorpan, any particular way you would suggest approaching it? Removing 100mm from the wheelbase and planning to take that out of the rear footwell.
It'll have a cage which will help overall but the floorpan, any particular way you would suggest approaching it? Removing 100mm from the wheelbase and planning to take that out of the rear footwell.
Great build, and basically what you said but with a Clio on top.CamMoreRon said:
I don't want this to come across the wrong way and make myself sound like a pompous tt, as I do appreciate questions, opinions, and feedback; but suspension design is my day job at an OEM and we don't even test to those extremes on commercial vans!
I have conservative enough predictions of 2g in braking / cornering and 4g in bump, and safety factors on top of that. There's no extreme customer use to be allowed for, so if I brake in to a massive pot-hole or smash a square kerb then I will expect something to get bent or broken - that's just life!
The FEA simulations I've done are pretty comprehensive, as is how I've calculated all the input forces & constraints, so I've got good confidence in them. I won't be going and immediately hammering up the hills.. as you rightly said I'll be sure to shake it down gradually and ramp up to "maximum death" after each check.
Whilst i agree with what you have written, with the best intentions, we are so far away from any OEM kind of activity here i don't think applying any stds from that realm is sensible. t, as I do appreciate questions, opinions, and feedback; but suspension design is my day job at an OEM and we don't even test to those extremes on commercial vans!I have conservative enough predictions of 2g in braking / cornering and 4g in bump, and safety factors on top of that. There's no extreme customer use to be allowed for, so if I brake in to a massive pot-hole or smash a square kerb then I will expect something to get bent or broken - that's just life!
The FEA simulations I've done are pretty comprehensive, as is how I've calculated all the input forces & constraints, so I've got good confidence in them. I won't be going and immediately hammering up the hills.. as you rightly said I'll be sure to shake it down gradually and ramp up to "maximum death" after each check.
For starters, a road vehicle will have probably half the lateral stiffness and 5x the compliance in it's suspension, and 10x the bump travel! One thing about race cars is that when you hit things they are so stiff the loads are absolutely enourmous, as i'm sure you are aware. (although, of course the "weakest link" is generally the wishbone / rose joint etc)
For your light car, taking the all up mass (600kg?) at say 2g with a 200% safety factor would be a good starting point would it not?
Also, i'd hate to try to accurately model your welding and it's affects on the parent material in enough detail to know what the stress distribution was like in those uprights? So, the better say safe than sorry response is just to hang a large known weight on em and at least pass an initial "idiot" test ;-)
In reality, if you couldn't hang say 2.4T off each and every attachment point on that upright i'd eat my hat, it "looks" easily strong enough. (if it is "stiff" enough, your FEA will probably be good enough, but getting a real, measured distortion figure (mm/kN etc) would still help to validate your model data)
Of course, with the car not being road legal, at least you have a much reduced liability to worry about for a pure race car etc!
Edited by anonymous-user on Tuesday 1st April 15:53
Well, aside from the debate on what loading is appropriate.. measuring any displacement is going to be incredibly difficult to do reliably. The loading on an upright isn't as simple as hanging x load off at an attachment point, because the part is restrained in different degrees of freedom at each one, and then each of those has it's own load that will need to be applied in a certain direction in order to be realistic. Then you have the problem of needing a frame stiff enough to constrain it (and knowing the stiffness of that to some accuracy) then a reliable way of measuring the applied force and the displacement - the order of which are very high and very very small respectively! Once you get in to it it quickly becomes an incredibly laborious - bordering on impossible - process to try and DIY.
Anywho.. it just isn't the way things are developed these days. In an OEM you only ever run FEA prior to a part being fitted to a vehicle; there's never any "hang a few tonnes off it and see if it breaks" about it, it's all done with modelling and then validated by not breaking on the prototype vehicle.
At this level, though, the best you can do is to make sensible predictions and accurate models, and have confidence that way. Otherwise you have to spend years iterating and re-checking, or over-engineer everything outright and have an overweight car - neither of those appeal to me!
Anywho.. it just isn't the way things are developed these days. In an OEM you only ever run FEA prior to a part being fitted to a vehicle; there's never any "hang a few tonnes off it and see if it breaks" about it, it's all done with modelling and then validated by not breaking on the prototype vehicle.
At this level, though, the best you can do is to make sensible predictions and accurate models, and have confidence that way. Otherwise you have to spend years iterating and re-checking, or over-engineer everything outright and have an overweight car - neither of those appeal to me!
CamMoreRon said:
Anywho.. it just isn't the way things are developed these days. In an OEM you only ever run FEA prior to a part being fitted to a vehicle; there's never any "hang a few tonnes off it and see if it breaks" about it, it's all done with modelling and then validated by not breaking on the prototype vehicle.
Indeed, and then we spend countless hrs doing FMEA's, risk assessments and 100 of hrs of shakedown/durability tests under controlled conditions on a close track before those parts are released anywhere near the real world! I'm going to guess you are not quite going to go to that level! ;-)All you need to do, is to clamp the wheel bearing spigot in a nice chunky steel jig (make it out of cheap I beam section from your local steel stockholder), put a dial gauge on the test point (say where the lower wishbone spherical bearing sits) and lean on it with a 2 ton bottle jack! if the jig breaks before the part, you're good to go, haha!
But in reality, as long as you have a proper methodology and program for your development activities once the car is built, and under reasonably controlled conditions, you'll be ok i think ;-)
CamMoreRon said:
Max_Torque said:
But in reality, as long as you have a proper methodology and program for your development activities once the car is built, and under reasonably controlled conditions, you'll be ok i think ;-)
Does "drive it and see if anything breaks" count? 
Yeah I've done stiffness analysis on the space-frame using simple elements, and proper stress analyses on all sub-components. The ROPS (roll structure) is all size-mandated so doesn't require anything special. As for determining bolt loads, I've used a comprehensive method with matrices to calculate all loads from the contact patch forces / moments. I've put over 2 years in total in to my chassis calculator, so I'm confident with the outputs.
I've ignored friction at all bolted joints and just spec'd fastener size with shear force. What do you mean when you say it has caused issues? Thanks for the book reference, I'll be sure to take a look!
I've ignored friction at all bolted joints and just spec'd fastener size with shear force. What do you mean when you say it has caused issues? Thanks for the book reference, I'll be sure to take a look!
Gassing Station | Readers' Cars | Top of Page | What's New | My Stuff