spaceframe bracing materials

[NeilRoshier]

I have an interest in trying ali skinned ali honeycomb panels as stiffeners in a simplified spaceframe structure. The panels are made from .5mm skins of 5052 Aluminium. The panel skins would be seperated from the core for 15mm around the edge of the panel and a 1mm 5052 square 'U' channel would be inserted and bonded to the skins and the core with either a structural acrylic or epoxy adhesive. This would then serve as a bond face at the edge of the panel and also as an internal local reinforcement for mechanical fixing. The panel would then be placed into the spaceframe and overlapping 1mm thick by 20mm wide 5052 would hold the panel to the spaceframe on both side of the panel, bonded with the same adhesive and mechanical fixing (probably SS rivets).
Does this sound a reasonable method of fixing such material to a spaceframe?

 

[GregLocock]

Yes, but it is about three times as complicated as I'd have used. Given that it is a shear panel you are trying to load the skins up, the core is a secondary consideration. Therefore, being nasty, I'd just glue the panels into the spaceframe direct, with no fancy edging. Your pop rivets are OK as a gluing fixture, but they won't take any load at all if the glue is working. The holes you make in the spaceframe might not be a good idea.

Choice of glue is interesting, you might look at polyester resin rather than epoxy, as it is stiffer, cheaper, and less exciting to work with. Put some microballoons in (wear a dust mask) if you want to keep the weight down, then you can go mad and splosh it everywhere (ie up into the core). You might even run glass tape along the join.





Cheers

Greg Locock

 

[NeilRoshier]

Yes greg it does seem a bit complicated but in some ways I do not mind that if it works better. Given the cost of the panels I would like to get as much out of them as possible! I had thought that getting the skin in double shear would have some advantages in some circumstances, specifically with ultimate failure in an accident (I do worry about such things and more importantly my wife worries too..showing that you are taking all care and time can make some life events easier in the long run!).
The glue is a reinforced structural adhesive that on the tech sheet is rated at 35mpa in shear with good peel qualities and cures at room temperatures. This was specifically recommended due to its ability to adhere well to both ali and steel and also isolate the metals from each other. The pop rivets were there just to limit possibility of peel in extreme circumstances, particularly in the cabin areas. I did not think that polyester made a good adhesive for metals?
I have had a look at the lola LM tub that used this material but they used extruded ali joiners between the panels. I feel that it would be easier getting a recognised 'structure' that is 'reinforced' with recognised materials past the scrutineers rather than a 'novel' structure. Also the 'spaceframe' is the recognised crash structure.

 

[GregLocock]

If you have a good source for glue info then run with their suggestions. Be very careful with surface prep, obviously. Will it need a post-cure bake to harden properly?

I was being diplomatic about the pop rivet holes in the spaceframe tubes, frankly, they are definitely not a good idea. The aerospace guys often recommend pop rivetting glued structures together for insurance, all they do is put stress raisers into the structure. Mind you, it does ensure good contact between the two parts as the glue sets. An option might be to weld a flange onto the spaceframe and glue (and rivet if you have to) to that, leaving the tubes unsullied. I'd discuss whether you need pop rivets with whoever is going to sign off the car (is this a track car or street?). If you need an SAE-A engineer to get a certificate for the road then I know one down here who understands this stuff somewhat (he's an ex solar car guy), and I can explain it to him. You might consider doing a test or two to see if your proposed structure makes sense - it is very difficult to get load into thin skins, typically they buckle along the edge of the reinforcement. While we are at it, is there really much performance gain over 1mm aluminium sheet?

I'm glad you are thinking about safety - it is the main reason I pulled out of solar cars. Cheers

Greg Locock

 

[NeilRoshier]

I guess I will know if there is a performance gain over 1mm ali sheet when I do a test panel. For some specific uses there would have to be a substantial advantage e.g. in the floor area. Given I was thinking of using the same thickness panels as the tubes 25/30mm the construction would be simplified. There is no standard thickness of the panels and not much added cost (or weight either) going for the thicker panels. The material itself is impressive in its stiffness and curiously I had considered its use in a solar car as it might have been possible to make a chassis from a panel?
The spaceframe was planned to be in mild steel, square section 1.2-1.6mm wall thickness (apart from mandated crash structure for CAMS purposes which is CDS round) in hopefully a simplified structure given the possibilities of the panels. This is a long term project so consulting an engineer at this stange may be a bit premature! (I am interested to find engineers who do this sort of work are who are interested in it) I have some cohorts in this strange idea, but the plan so far only goes to the Power/weight ratio. Plan was for wet weight of approx 800kg and 180-200kW...but everything else is fluid, (and damn that gordon murray bloke...we can't get over the idea that the cental driver position is a good idea!). The vehicle is intended for road use at some stage.
The rivets are as you note, not the best idea in some ways but do seem accepted practice and they do also, as you note, hold things together as the adhesive sets. The clubmans have used the brazed on flange that you describe, but when talking to some of the constructors they have noted no difference in service or when tested (too low a stress perhaps?). There is a defined preperation schedule for the different adhesives and all adhesives investigated cure at room tempreture (I can access a large autoclave...but can't afford the rental...same cost as a good pair of koni 8212 dampers!).
I guess I had hoped to limit the possibility of buckling of the skins by the 'staggered' reinforcement of the skins when in double shear (ie say 10mm one side and 20mm the next) but do not know if this is a good idea?

 

[GregLocock]

I agree the floor is an obvious area to use this stuff, having said that you might consider going to composites - the same construction in glass and Nomex was our basic structural material for large parts of the body.

I suggest you look at feathering the edges of your reinforcement if you get problems with buckling of the skins. A 20 degree chamfer is none too fine, we used about 10 degrees. Cheers

Greg Locock

 

[NeilRoshier]

I had considered a full composite monocoque since I drive one everyday already. Only problem was getting it past ADR's and engineer approval, my car is too old to need to worry about such things. I was convinced to go the conventional route after talking to cambell bolwell, he should know! The composite panels are available and I had also looked at Klegcell to layup over. However the aluminium has some advantages such as a predictable response to accidents and ease of replacement for repair, as well as a few others.
As you know you do not design the chassis first, but knowing the properties of a material given the technology of the user to implements its use will govern some design choices. The design will truely start when we decide to go front/mid or mid/rear and also when we find a good front upright/brake combination that is easily available (after all there may be more than one car in time). The rest of the suspension will be detirmined around that point I think.

 

[NeilRoshier]

I will be making the rectangular test panels in the next few weeks and will do 3. One frame with no bracing, one with bracing and one with ali panel bracing attached as described above. Testing will be very simple. If I clamp the frame to a bench at one end, support one of the unfixed corners and then progressively add weight to the other corner thus measuring the 'twist' in the panel with a DTI. Does this sound reasonable Greg?
Frames will be mild steel, brazed together (as per clubman chassis) with ali sheet reinforcement and ali panel both bonded and riveted.

 

[GregLocock]

Yes, but you are measuring the torsional stiffness, whereas by its nature you are really mor einterested in the shear properties.

ASCII CAD: Side view

X-----------------
| |
| |
X_________________|
|
\/
F

Bolt it to the side of your enemy's house at X, then apply weights at F






Cheers

Greg Locock