CR14 Frame Design

[kabhi]

Hello everyone,
I am kumar from Team Camber Racing, SRM University. I am currently working as Chassis Lead. We are going to participate in FSG for the first time in 2015.
We have participated in 'SUPRA-SAEINDIA' last year. Below is the image of the frame of our old car.
Please provide your feedback about the frame and various areas of improvement in it. It will be helpful while designing our next car.
CR14_ses.jpg
CR14.JPG

Thanks.

[Adam Farabaugh]

What are your goals for the next one?

Few things I see right away:
No shoulder harness bar triangulation
No triangulation on motor mount points
Jacking bar would be loaded totally in bending
Potentially overbuilt, not sure why you have thick walls everywhere

[Charles Kaneb]

Take a look at the thing from a top view. You'll see lots of rectangles. Rectangles are just parallelograms that have 90 degree angles on the drawings. Even without an input load (control arm, rocker, steering column) that pushes in a way that would deform them, they'll come out well away from 90 degrees just on the welding table.

I also think this frame would be painfully difficult to weld - ask yourself: how will I get the X-reinforcement into the box below the driver's seat? how will I get the x-axis braces in the box in front of them? "Grind it till it fits and weld it till it fills" is suitable for repairs at the kart track with spare rollcage tube and a coathanger for filler wire when you've broken something, not for something whose design and construction are getting weeks of analysis.

[MCoach]

I'd say there are way too many tubes. Too heavy. One thing that I learned greatly and wish to pass on is that you never need as much stiffness as you think you do. Floppy frames are fine as long as they are consistent. Just aim for some frequency in torsion (7th free body mode?) that is at least a decade higher than your roll stiffnesses. Inherently well designed frames will bring the vertical and horizontal bending stiffnesses along with them.

The x braces seem fine from my view. I've done them like that before and it didn't require any extra work to do so. Fit the first tube across the span and split the other one into two pieces which become easy enough to fit and fall into place with precise tube coping/fitting. As long as you plan out the fitment of the tubes as you would apply them one-by-one, then it becomes fairly straight forward. Don't think of it as what the easiest sequential fits will be but more so the most rigid and accurate fits first, then just deal with the complicated fitting notches later on. By the time you get to difficult ones it will seem trivial after the first few fits.

The rear structure seems overly complicated. I'm curious as to why you have 3 horizontal tubes in the rear? There may be more mounts than you need back there and definitely more than you need for suspension mounting. I wouldn't worry about the jacking bar as that isn't loaded very heavily or often. It could probably be like 12.7mm or 18mm tubing for the supports.

The front looks about right but your diagonal tube does not terminate at the bottom of the roll hoop as required. Beyond that, just another comment on thinner wall and smaller tubing should be used for all non-required tubes.

The side impact just caught my eye, make sure the upper side impact tube actually meets the height requirements over it's full length. Right now it looks as though it may not.

The main hoop has quite the flat surface on top and can be more "triangular" to shed some weight. Which brings up another thought. I'm not sure what you're packaging requirements are beyond Percy and other templates but it looks quite long. Just a thought, but not a criticism.

Overall, without much data to go off of, I'm impressed. Good work.

Could we get some numbers on projected weight and stiffness simulation (or target...or both) values? We can make this thing hella cool. Good luck!

[kabhi]

Thank you MCoach and others for your valuable comments.
First of all I would like to mention the reason for selecting the tube dimension greater than minimum. All the tubes available were of 25.4 mm, but as you know there are always tolerances (even while manufacturing from supplier), I thought to be on a safer side by selecting tube of higher OD and wall thickness. I have refered a few post on forum where it is mentioned that about -+10% is allowable, but I didn't find anything in rulebook of this sort. It will be really helpful if you can help me out on this. This would save 8kg on our current frame straight away.
@Charles as MCoach has mentioned it was no problem doing X-reinforcement as we fishmouthed all the tubes (Used SolidWorks for that). But yes we faced problem while welding the A-cabin joint. I will definetely work on changing basic frame shape to a more of triangular type.
@MCoach, actually we had differential mounted on the rear bulk head. So in order to provide enough strength to the bottom two nodes of bulk head I traingulated it. Can you explain me in detail which member you meant when you said ''The front looks about right but your diagonal tube does not terminate at the bottom of the roll hoop as required" ? About side impact member I assumed the centerline of tube to be in 300-350mm range from ground (Isn't that right?). And yes we are going to change the top of main roll hoop as you said.
Few datas are given:
Chassis weight: in software:40kg(without suspension mounts), actual: approx. 50kg(with weld and suspension mounts)
Torsional stiffness: 1500 Nm/deg
We are targeting to reduce the overall weight to be around 30kg including suspension mounts. In order to do this we have planned to go for square tubing for the members of suspension hard points. This will allow us to go for machined Aluminium suspension mounts (it was made of mild steel plate of 5mm) which can be bolted to square member. Also we can go for minimum tube dimension and can reduce the overall length of chassis.
Sorry for writing such a huge post.
Thanks again for your comments. Waiting for more feedback.
I will find more pics of it to post here.

[kabhi]

Here is link to 3D PDF file of frame which can be viewed straight away in Adobe Reader.
https://www.dropbox.com/s/wvg6bq7m7a...assis.PDF?dl=0
I am also attaching the image of our car in its initial phase:
CR14 _2.jpg

[Jonny Rochester]

To me the chassis looks good. You can reduce the number of tubes and mike it lighter if you think hard about it, but it depends to what level the rest of that car is built to make it worthwhile.

According to rules, you don't need harness bar triangulation, as it is straight and not bent. That is OK.

According to my strictest reading of the rules, the lower 180 degrees of the jacking bar needs to be available. As you have a support tube interfering with this lower surface I'm not 100% sure that is OK, and if made me add a new jackingbar to our chassis. Does anyone else have an interpretation of this?

[Adam Farabaugh]

According to rules, you don't need harness bar triangulation, as it is straight and not bent. That is OK.

My point wasn't that the harness bar is bent. But the nodes where the harness bar meet the main roll hoop are not triangulated. So torsion isn't a problem, but in a crash situation buckling of the main roll hoop due to driver inertia is. Remember that the driver is the heaviest part of the car. Regardless of what the rules say this is not a good engineering design decision.

[kabhi]

@Adam we have run the tests along with the forces as specified in rulebook. Doesn't it cover this scenario?
Actually we are going to change harness bar this time. We are planning to make it bent and hence that needs to be triangulated.

[kabhi]

All the tubes available were of 25.4 mm, but as you know there are always tolerances (even while manufacturing from supplier), I thought to be on a safer side by selecting tube of higher OD and wall thickness. I have refered a few post on forum where it is mentioned that about -+10% is allowable, but I didn't find anything in rulebook of this sort. It will be really helpful if you can help me out on this. This would save 8kg on our current frame straight away.

Any replies form teams who have participated ???