my project car!

[rockit]

hi all,

thought it was about time to share my current project!

the general idea is: a mid-engined, light weight, road legal, AWD car using the driveline from a VR4.

its still very much in the design phase, nothing has been approved at this stage (although i have discussed some of it with a engineer, but its still very much theoretical until i can hammer down the details more)


a bit more specific detail:

Chassis: hybrid tubular spaceframe, monocoque, rail chassis.

spaceframes make up the front and rear subframes, a monocoque forms the passenger compartment, (there will be no roof, although i intend on adding one, but it will not be structural) , with 2 X braced rails running along the outsides of the passenger compartment for extra strength. My original intention was to go pure spaceframe, however that made adding doors somewhat difficult, so a rail/monocoque was devised to allow doors to be fitted.

my preliminary CAD modelling has put the chassis rigidity and beaming strength well over that required.


Driveline: complete VR4 driveline (excluding AYC stuff, and transfer case)

the engine will be placed in the rear of the vehicle, the transfer case will be replaced with one from a EVO, which will allow it to be rotated 180 degrees to output under the engine. with the standard diff (minus AYC) at the front of the car. Note: i have yet to purchase a EVO T/C to make 100% sure it clears, but will be doing so soon. this is currently the biggest issue that i have with my design, as obviously if it doesnt clear then there goes the whole idea!

the rest of the driveline will be standard. (may require the propshaft to be shortened slightly)


Suspension and steering:

i have yet to confim any details in this area (if you ever try to design your own suspension geometry you will find out how confusing it all is!). but i will attempt to re-use as many components from the VR4 as possible.
i imagine i will end up handballing this to the engineer, as my understanding of vehicle handling characteristings is limited, and its vital that this part is done well.
obviously major changes here will result in changes to the chassis, however my simulations have found that minor alterations wil not massively compromise the chassis strenght.


Body:
I will use a fibreglass lay-up over shaped foam for the majority of the bodywork. I have yet to confirm a body shape, but as it is the last thing that actually gets done then i still have time to think about it!


Obviously I have yet to hammer out some details, and I have made some fairy crucial assumptions here!

some assumptions i have made thuis far: please correct me if im wrong!!

1) that the front driveshaft spline count matches that of the rear driveshafts (ie I can put the engine in the back and still use the front driveshafts to drive the back wheels)
2) that the rear driveshafts have CV joints that can accomodate steering input.
3)that ABS sensor is the same back and front (as above, although not crucial as ABS may be disabled anyway)
4) that the evo T/c will clear under the engine
5) that it will meet emissions standards


obviously i havent had access to all the details i would like, so most details are fairly vague at this point, however I have most of the vital details that I need in order to seek approval, and after its approved then i can fill in any gaps.


here are some pics

the first is a pic of the chassis, its a few revisions old, since i rendered that image I have shortened the front, redone the beams in the front (so when a frontal impact occurs the driver no longer gets a tube through his/her heart!), ive also added more crossbracing, and re-done the suspension to more closely resemble that of the final design. obviously there is alot missing from this picture (the engine for example), these have since been added into my model, but i havent rendered any images since then.

any input (or help) is appreciated.


It should be noted that Im an engineering student, I am doing this project as a learning experience (this is why I am doing most of the design myself rather than letting the engineer do it). my design in the end must be approved by a charted engineer (he has to check my work and then issue an engineers report on it) and then by Transport SA.

 

[JUSTCRUZN]

that last pic of the soon to be car looks wicked mate. im excited to see it when its finished. so how long you got to make build this thing? just interested sounds like it will be a fun project for you

 

[rockit]

well I want it to be done inside a year, thats from starting the ball rolling (officially) to the car driving legally.. im not sure if this is being pessimistic, or optemistic! as various people have said between 6 months-5 years!

oh i should have said that in the final pic the W in the front grille wont be there! that was part of the chassis sticking through the bodywork (as i hid the chassis when working on the body) the chassis has since been re-designed to reduce weight, and is now well clear of here.

also my rendering skills are somewhat lacking at the moment, there should have been various textures (like the front grille) but the program im using appears to have lost them somewhere :s

 

[unclepaulie]

few questions:

will the engine be mounted east/west or north/south?

why choose a 6a13? (i hav reasons why not to choose such a motor, even though they are good)

just quickly examining, that current frame wont suit your chosen engine/driveline layout. currently its akin to a mx5, so central mounted cab. preferaby you will need a frame similar to your ferrari's and lambo's which hav a cab forward design/longer rear section.



you mention your knowledge of chassis design is limited, i dont see a cheap way of transfering the vr4 running gear into a tube frame and switching the engine location/mount position

if you were to put the motor/gearbox combo straight in, it would be much easier. this way you could design a frame like a caterham and have the seating position mounted further back.

 

[rockit]

thanks for the input..

yes i realised the passenger cab was all too far back when i put a engine into the assembly, it has been revised since then (as you said the passenger compartment had to be shifted forwards)


the engine will be mounted E/W (the same orientation it is in a VR4)..

as for the engine
honestly i dont know why i chose a 6a13, i think it started when i saw one cheap on e-bay near the beginning of the year, and it made sense at the time (one of the major reasons is that it can be converted into mid-engined easily and still be AWD (hopefully), with minimal changes to the rest of the driveline! if you have a better alternative in a similar price range then let me know and ill investigate

 

[rockit]

hey, just saw you edidted it while i was writing my reply lol..anyway im aware that changing the engine location will not be cheap (or easy)

as I said i will do this by swapping our T/case with one from an evo (the output shaft is on the TOP of an evo T/case) this will then allow it to be rotated so the output feeds underneath the engine (points to the front of the car), from there its just the rest of the VR4 running gear.

If i run into any insurmountable problems i will simply do as you say and drop the driveline in as it is in a VR4, but i want to leave that as a backup plan at this stage..

 

[PeteGT]

Looks like a hell of a project there Rob. Can I ask what uni your are doing your degree at? Also what year are you in?
If you're at UniSA or Adelaide I'd suggest going down to the Formula SAE team and having a chat to their chassis guy/gal. You'll learn a lot if they have the time to speak to you.

The main questions are on your goals for the vehicle:
Do you want to race it or just use it on the road?
What is more important to you, structural/performance or easy to build and low cost?
Where are you getting your load cases from? What are your structural goals and why? (torsional strength etc.)

Some constructive criticism:
First off; Load Paths. You ideally want to have your loads applied and reacted at node points, where 2/3 or more tubes come together. This is the basis of space frame design, using 2d/3d triangles to make a structure. Where possible avoid open square panels, triangulation is key. Have a think about where the loads will be generated and how you will be reacting them. You will want your inner wishbone points to be at nodes, or damn close to them. Also think about your top spring/damper mount structure. You will want to react it as efficiently as possible to the bulkhead.
It looks like you have some likely bulkhead structures forming up at the front and rear of the cab. Get the suspension reacting into these and you're well on the way to a decent frame.

Is there any reason for you not wanting a structural roof? It will cut down the amount of bracing you need through the cabin, improving your door access too. The cabin you have at the moment looks very flimsy.

What are you planning on making the cabin from? If it's composite I wish you luck, it's going to take you a while. You could get away with a space frame cabin box pretty easily. If you really want to go roofless you can run some triangulated box structures/rails and have a high door sill line. (Merc gullwing anyone?). If a roof is acceptable a spaceframe gets even easier. Take a look at a rollcage triangulated to the suspension in a modern touring/GT car for ideas.

How were you intending to build it? Do you have ready access to machine shop and fabrication facilities? Spaceframes are VERY time consuming to build, if possible design in flat planes that can be cut/welded together on a jig board and then assembled together.

Some reading for you if you havn't done so already.
Search 'locost sports cars', they're not exactly what your project is but they're quite similar.

Race Car Chassis Design and Construction; Forbes Aird. Goes over a lot of the basics of chassis/race car design. Good examples of do's and dont's.
http://books.google.com.au/books?id...s aird chassis design&source=gbs_similarbooks

Tune To Win series by Carroll Smith. These books are invaluable (and only $140 delivered) if you want to understand the principles of your project. They cover not only the structure/suspension part of the car, but also all the ancillaries like plumbing/throttle/brake etc. A great resource for your degree too, lots of down to earth engineering.

Heaps more books around on chassis/race vehicle design. Google is your friend.

Bit of a long winded reply but there is a lot for you to understand. Are you doing this as part of a design project? Might be worth seeing if you can get some extra course credit for all this work. If I can be honest I think you may be struggling to get it on the road and legalised in a year while doing Uni at the same time.

Don't know what background your engineer friend has but sometimes it is handy to get feedback from someone who has done a similar project before. If you want to ask any detail questions my email is pete2good@gmail.com

Hope I haven't scared you off.
Good luck man

 

[rockit]

Looks like a hell of a project there Rob. Can I ask what uni your are doing your degree at? Also what year are you in?

Im at Adelaide uni, and next year is my final year. Im doing an Aerospace degree though so alot of this is new ground to me! I will try to get in contact with the FSAE guys, hadnt thought about doing that.

Do you want to race it or just use it on the road?
What is more important to you, structural/performance or easy to build and low cost?
Where are you getting your load cases from? What are your structural goals and why? (torsional strength etc.)

mainly road use.
structural/performance is more important but cost is also an issue (obviously)

You will want your inner wishbone points to be at nodes, or damn close to them

I realised this quite quickly actually (there was more deflection in the beam supporting the wishbones than there was in the entire rest of the chassis), this has been corrected in my updated chassis. I will have to get a picture of my most recent revision as I hadnt realised exactly how much i have changed since that image was taken!

Is there any reason for you not wanting a structural roof? It will cut down the amount of bracing you need through the cabin, improving your door access too. The cabin you have at the moment looks very flimsy.

Ironically that was the engineer i was talking to who said to remove the roof! I assume he has a reason.. i personally wanted a structural roof, but in the end he has the final word on approving it :s.. and your right the cabin is obviously the most flimsy part on the chassis, however due to the large rails, cross bracing (not shown in those images)and the transmission tunnel it is still well above the required torsional rigidity.

What are you planning on making the cabin from?

sheet steel. Im avoiding structural composites as they would be hard to fabricate, and would require another test to get approval.

f you really want to go roofless you can run some triangulated box structures/rails and have a high door sill line.

In doing research I found an analysis that indicated that a triangulated sill structure could be replaced with an acceptably large tube without severely compromising integrity, sure its no longer a true spaceframe, but almost no car really is!

one of my first designs did have a triangulated sill structure (similar to what was in the lambourghini countach i believe) but I found that the tube was almost as rigid and obviously simplifies it greatly

How were you intending to build it? Do you have ready access to machine shop and fabrication facilities? Spaceframes are VERY time consuming to build, if possible design in flat planes that can be cut/welded together on a jig board and then assembled together.

I have access to most of the tools I would need, I own a milling machine, lathe and a good MIG welder (and all the normal tools), any I dont have I intend on buying along the way. and have access to a manufacturing workshop that can make me parts if I cant do them myself.



I have already searched for locost cars, unfortunatly many of those sites dont go into details like calculating torsional rigidity, suspension geometry. But I still have found them useful to see how something like this all goes together.

I intend on reading up on books as soon as uni is done. also will actually officially start it going, which means 'officially' consulting a engineer, hopefully they will help me with anythign im having difficulty with!



thanks for the input, and no you havent scared me off, i was really hoping someone would come along who knows what they are doing and can help me out!

have you been involved in a project like this?

 

[bradc]

The EVO 4-6 transfer case is the same as a VR-4. The EVO 7-9 is a different shape, are you planning on using one of these?

 

[mookers]

If this car ends up being successfully built, you have to drive it to Melbourne for a wash meet lol

 

[PeteGT]

I'm interested to see your updated design, can you put up a picture of it? Screen shots are fine.

There was a US based FSAE car a little while back that used carbon tubes running along the length as the primary structure. These guys I think:http://dot.etec.wwu.edu/fsae/viking30.htm.
From a purely structural point of view it's a good solution, just pay a lot of attention to the end fixings.
Your uni should have CAD packages with FEA solvers available to use, just be careful with how your loads and constraints are applied, they'll have a big effect on your results.

I do have a bit of experience in fsae with Monash. Designed the 08 chassis, and built the 2 years previous.

 

[unclepaulie]

+1 for updated chassis pics

 

[Guest]

http://www.carbontech.com.au/

I would strongly suggest trying to have a chat with the guy that designed the Redback Spyder.
He's doing a MkII version now, but his first version was pretty much the same as what you are doing now.

He wanted to build 'his' dream car the way he wanted, and to have it legal to drive on the street.
He realised that at the end of the day he could re-coup a lot of his losses by marketing the vehicle out to the general public.
from what i remember of it, it has been quite large for him in the US.
His prototype ended up costing him something like $1.2 Million? from splooge in his pants right up to turn key ready for public sale and export.
He still hasn't got it as an Australian complianced vehicle, but its in the works.

He needed to build and engineer the car as an ICV (Individually Constructed Vehicle) and then re-classify it for small volume production.
To comply as an ICV, it needed to meet or exceed ALL of the current ADR's applicable at the time of testing.
you really need to know what ADR's you are going to need to comply with.
It may need ABS, ESP, Airbags, compliance with pedestrian impact rules and a whole lot of other bits and peices, ect ect ect.
It may work out a lot cheaper and easier to build it for track use only, maybe you can get club rego as part of a race club? Or special rego for Targa races?

I would LOVE to build something like this, and one day i hope to start on something, It'll probably be a custom bike or trike though, just coz thats a little more down my alley (20B turbo drag trike?)
I really hope you keep at it, dont lose sight of your end goals.

Anyways, GL with it.

 

[rockit]

hey all, have been busy with uni work all day, but will get updated pictures as soon as i can .. im trying not to open the CAD program right now, as everytime i do that i see soemthing i need to tweak, and before i know it hours have dissapeared!...

The EVO 4-6 transfer case is the same as a VR-4. The EVO 7-9 is a different shape, are you planning on using one of these?​

yes i will be using one out of a 7-9, trying to source a cheap USDM one which didnt come with ACD... this will require modification in some form to bolt onto the gearbox, but im sure it can be achieved

Your uni should have CAD packages with FEA solvers available to use, just be careful with how your loads and constraints are applied, they'll have a big effect on your results.

Im currently using Solidworks 2010, using its built in FEA package. I have fixed beams mounted to the wheel hub assembly front and back. the back beam is fixed in place, while the front beam has a 6kNm torque applied to it (3kN up at one end and 3kN down at the other end of the 2m long beam), aswell as a sliding pivot at the center. these constraints mimic the actual test setup stipulated in the ICV booklet (although I will need to ask how the engineer does his testing, as various websites show different setups so that may be up to the individual engineers and waht they prefer).

If you know of more appropriate load/constraints then let me know and ill try them aswell.

To comply as an ICV, it needed to meet or exceed ALL of the current ADR's applicable at the time of testing.

I believe SA is the only state at which doesnt require this. in SA the requirements are frozen from the time that approval has been granted (at least this is what the engineer I talked with told me!). ie. if i get approval this year I can take 20 years to build it and i will still only need to comply with the 2010 ADRs.

Also several ADRs have been excluded from the requirements (such as the ones requiring crash tests). I believe Airbag and ABS (and ESC) requirements are also excluded (but i will need to check this!), as it is very likely that a incorrectly implemented ABS or airbag system could massively decrease the safety of a ICV.
And they are also fairly lenient with the emission regulations (ADR 37/01 is accepted rather than the much more stict current regulations, and ontop of that they have other alternative criteria!)

It may work out a lot cheaper and easier to build it for track use only, maybe you can get club rego as part of a race club? Or special rego for Targa races?

one of the major requirements I set was that it would be road legal, and so im not going to settle for anything less. yes it undoubtably makes it more expensive (have been quoted $5k for engineering fees/tests alone (2k+ for a single report!!)) but in the end i would much rather have a vehicle i can drive on the roads than one that cant be.

 

[Fully]

I think you will run into trouble, Not sure of your model scale ( Looking at tube gauge), doesnt look strong enough for a street car.
The torque will make it flex.
Box design limits what you can do.

Why dont you start with a chassis like this?

 

[rockit]

those chassis are surprisingly weak, they are simple rail chassis without any form of X bracing at all.. possibly some of the least rigid designs possible..

the tube size im using is 1" (well 25mm)x 1.6mm and this may surprise you but in most of the chassis it will be able to be replaced with 20 mm without severely impacting the rigidity.

I doubt i would be able to modify a simple rail chassis to pass the required rigidity.. althoguh once a passenger cab is bonded onto it it *may* pass.. (as the passenger cab would add alot to the strenght)

 

[Tom]

This project is going to be AWESOME!! I want to follow it =]

 

[smohekey]

I've been planning something almost exactly the same as this, however I have a quite different idea on how to achieve 4wd.. here's a hint, FWD manual transmission on a North/South oriented engine

 

[unclepaulie]

last thing you want to use is a twin parallel rail chassis.

id go with tube frame. triangulate everything. and hav the suspension and motor mounts located on a node of a few triangles.

that way you can have some decent load dispersion.

 

[rockit]

here's a hint, FWD manual transmission on a North/South oriented engine

with propshafts going to front/rear differentials?...
I thought about this, and still may use it if i cant get this way to work.. but i think theres a issue with gear ratios, as the FWD gearbox has the final drive ratio built into the 'front' (now center) diff, but the diffs will also have their own final drive ratios aswell.. so you will end up with something like a 16:1 final drive (corect me if im wrong, as that was one of the major drawbacks i found to that design)..



ok got some updated images...

updated chassis, can still see a few things I need to improve (re-position the rear nodes so they actually sit on the suspension points), also add some Y bracing along the top of the engine bay. also the bar across the front suspension seems to be overkill, so will probably replace it with a lighter alternative.. also need to shift the front lower A-arm mount points to be closer to the nodes, as one is ~30mm off and thats producing a huge stress(well compared to everything else, still well off failure) in the beam its mounted to

heres a basic stress plot, i have ignored all suspension components (as the simulation takes about 4 hours to run with supension members and their contact conditions)

I have simply replaced the suspension with a large brace that locks into the suspension points, obviously its not completely accurate, but the results are similar to those with suspension included and takes much less time to run, so gives me a good idea if any modifications i make are heading in the right direction. Also im still not 100% sure if suspension is included in the final test, as from research around the web it seems that most engineers dont use the suspension points at all, simply attach a large beam under the front of the chassis and apply a load to that!

I should note that in the image above the deformations are scaled by 140 times, otherwise there is almost no visible deflections. also the ghost image is the undeformed object.