It will be interesting to see the outcome of these analyses in that how much a sunroof effects effects shear and torsional strenght of a unibody.

However it remains without a doubt that in a monocoque or stressed skin structure that a modern unibody design essentially is, that putting holes in the stressed skin will make the latter less effective and negatively affect shear strength and concomitantly overall torsional rigidity. That is closed sections have greater shear strength (that's why the bottom of the can is stronger).

Some relevant readings for those interested in that regard are:

P Kuhn: Stresses in aircraft shell structures.

RM Rivello: Theory and analysis of flight structures.

ME Heerschaap: Design of cut-outs in shell structures.

 

Once again Im trying to tell you the modern automobile is not a good example of a stressed skin structure. Airplanes, yes. Some race car designs, yes. Everyday cars, not really. If the automobile was a stressed skin than the roof would be made of much more than sheetmetal, and roofs would not dent as easily inward.....(hail, drunk guy on the roof, etc...)

 

It's a stressed skin - the strength is in tension, no other mode. It has no need of puncture resistance, as far as its primary function is concerned. The roof can be pressed from tissue paper so long as the underlying formers ensure it retains shape. Indeed earlier aircraft were little more than doped paper stretched across wooden formers, and in the case of the Mosquito for example, that provided all the rigidity neccessary to handle a couple of Merlins.

Car bodyshells are an excellent example of a horribly compromised stressed skin monocoque - that's why they're so heavy. If they didn't have so many holes cut in them they'd be half the weight. Of course, you wouldn't be able to see out (assuming you could get in), which might create problems in traffic :-)

Tubing. How stiff is a hollow tube ? Stiff. Better question - how stiff is a hollow tube when you weld a plate over each end ? Much, much stiffer.

The squeaking sunroof is a great example - my 98 had that from about 6 months old. I have a hollow at the end of my drive into which a rear wheel will drop when I turn around. That squeak is shell flex...

References, those I can remember off the top of my head :
Colin Chapman's book on racecar design, Mike Costin's book on same, Carrol Smith's books, Allan Stanniforth's books, The Major - Arthur Mallock, lots of stumbled-upon SAE papers); general structural - Grand Prix Motorcycle Development, Tony Foale's books. One book I happen to have next to me - Fundamentals of Vehicle Dynamics, Thomas D. Gillespie, though actually only touches on the subject, as it pertains to geometry control. Apologies to those I've omitted

Build a square, fix a side to a table, pull and push on one corner. From the same material build a triangle, pull and push again.
Add a diagonal to the first square - see how much harder it is to deform ? Remove the diagonal, staple a newspaper across the square (lots of staples though) - now see, just how dramatically stiffer the square is ? Now cut a big square hole in the centre of the paper, see what force it takes before the paper tears. I give this exercise because it is the one given to me that opened my eyes.


Charles

 

Hi Charles, I'm not disagreeing with you, but have a question. Is it possible that the squeeking you and Kayaalp are referencing could be related to travel in the roof glass and not the body deforming? The set-up used to hold most retractable roof glass in place doesn't seem to be all that sturdy.

The only times I've had my sunroof squeek on my Evo involved sharp, uneven bumps. When I put some presure on the sunroof with my hand, I found it easy to make the glass move. It certianly is not mounted well enough to resist strong vibration.

 

Bear in mind that my squeaking is from a DSM, not an Evo. The depression at the end of my driveway has sloping sides, my rear wheel is dropping into it at barely walking pace. As the wheel rolls into it, my sunroof give a slow squeak/groan until the wheel reaches the lowest point; same coming out of it. Also, it almost completely disappeared after putting a rear strut bar in the car.

I agree that with a short, sharp shock, the glass could bounce, thus creating the squeak, but with a slow progressive loading, it will be shell flex.

I would expect an Evo shell to be considerably stiffer than a DSM - not least of the reasons being the rear bulkhead and the lack of a hatch. However, after you add in race tyres and appropriate supporting mods the stress the shell is under rises dramatically, and may well bring you to the point of a sunroofed DSM on street tyres.

Cutting a big hole in the roof is going to hurt structural integrity no matter what, the question becomes whether the intended use will create loads that will reach that critical point. If they do not, then other than the poorly placed weight, the sunroof doesn't matter. If the loads do reach that point...

Here's an interesting observation, if one cares about such things. A few years ago I noticed ripples developing on the rear quarter panels of one of the quickest autocross DSMs around. At the time they were attributed to something else. Wandering around my car last season I started to see the same ripples developing in the same places. Stress. The funny part - to me - about this is that this was a tell-tale sign of a well-beaten-up Ford Escort race car back in England 20 years ago. Some things never change I guess. I should ask DG if he's noticed any on his car; he probably runs a full roll cage these days.

Charles

 

I understand stress and deformation; we scrapped the unibody of our ice racer every spring. I guess what I'm saying is that the Evo seems to be far more resistant load deformation than almost any other car I've owned. I think the only other car that had a unibody that felt this strong was my old 930. The bits attached to the Evo's unibody don't have that same solid feel, though.