Aluminum question

Date 08-03-2003 07:27

The only way to know for sure is to weld it. It may crack, and it may not.

As a rule, however, autogenenous welds in aluminum will crack at some point.

Good luck

brande

By MikeR

Date 08-03-2003 08:07

How 'bout this... It's a piece of round tube (6061) going through .100" 3003 sheet. The tube will end flush with the sheet and needs to be welded there for sealing purposes. On the bottom side of the sheet, I can't weld all the way around the tube, but what if I welded as much as I could from the bottom, then fused the top around the tube? You think it would have a better chance to seal (not crack) with the support of a fillet on the other side?

By CHGuilford (****)

Date 08-03-2003 13:30

Everything I have on aluminum welding recommends using a 4043 or sometimes a 4047 filler metal for 3003 to 6061. I can see why you would like to weld it autogenously given your application. Like Brande, I'd say give it a try. If it doesn't work you'll know and can go to the fillet and/or use 4043 on yourseal weld.
Would you let us know whether it did or did not work out for you? Many of us would appreciate knowing.

Chet Guilford

By MikeR

Date 08-03-2003 17:19

So my next question would be, how long should I wait until I call it a success? I mean, if it doesn't hot crack right away, what then? What I'm building is a set of valve covers for a customer, and he wants the bolt tubes to be flush with the body of the valve cover. Normally I extend them up a bit so I can put a fillet on 'em. How do I tell this guy that I don't know if it's going to work and not sound like an idiot? I don't know how comfortable he's going to be plunkin' down over $400 for these and not know if they're even going to seal. They are going on a dry sump motor (which pulls a vacuum in the crankcase) so they HAVE to seal.

By flatjwl (*)

Date 08-03-2003 19:49

You will know almost immediately. In a similar situation I have chamfered the side of the tube which meets the sheet which will allow some 4043 filler to be used without excess accumulation to stand above the surface. I assume you are using 1/8 thickness 3003?

By MikeR

Date 08-03-2003 20:44

.100" sheet to 1/2" x .065" 6061 round tube. Awfull small area to be doing what you suggested, but worth a shot! Thanks.

By flatjwl (*)

Date 08-04-2003 03:36

Yes, I agree, that is a thin wall. I was expecting a thicker wall tube. We used thicker (.120) wall 6061 and reamed after welding.

By awill4wd (**)

Date 08-04-2003 11:49

Mike, coincidently I also build custom valve covers for customers here in Australia as an after hours interest.
For the racers who are after weight savings I use 1.6mm (.065") 5052 sheet and for the bolt tubes I use 5/8" solid bar (6060 or 6061) which I drill 1/4" for a neat fit on initial setup. This gives me a nice solid tube to weld to, approx 3/16" wall thickness.
I use a 6mm thick flange bolted to the head and build the valve covers on the head. On my 1st prototypes I was finding the bolt tubes moving significantly even though they were bolted securely during the welding process.
I found the remedy I needed was to build a steel girdle arrangement which I then bolt above the valve cover using spacers to allow welding without removing the girdle until completely welded. I use 1/4" UNC bolts by 6" long so I can get the girdle up pretty high to allow good access for the tig torch.
The girdle a made allows me to build Chev small block, Ford Windsor and Holden (Australian V8) valve covers. One of my next projects is to build a girdle for big block Chev/Ford and Ford Cleveland
When fully welded I drill the bolt tubes out to 9/32" and spot face so they are square to the base. I then drill the bolt tubes 3/8" diameter just on the top about 1/4" to 5/16" deep. This allows me to flush mount 1/4" unbrako's to bolt the valve cover down to the head. I also use an O ring around the bolt to provide a positive seal.
Personally with welding, I don't like 3000 series sheet (too soft) or 4000 series filler as it doesn't colour match with anodising (welds show up black) For me, I prefer 5052/ 5251 sheet and 5356 filler.
Sorry for the long post, Regards Andrew.

By MikeR

Date 08-04-2003 22:10

Andrew, sounds like some nice valve covers to me. Out of curiosity, how much are you selling those for? I don't know what the exchange rate is like, but I'm doing this set of BBC valve covers for around $400. Yours must be much more than that with all the work you're doing. But probably worth it! How do you do the flanges for your valve covers? I thought about having them CNC'd, but for cost's sake I have them burned out on either a water jet or CNC plas table. (using 3/16" plate)

The reason I use 3003 is because I also do custom "water" tanks (wink wink) and the ductility the 3003 offers is essential for crash or off road scenario's. I just don't want to stock 2 different alloy sheets in several thickness's. My shop is too small for that. Also, I normally use .063", but this guy is going to have them engraved, thus the .100".

By Niekie3 (***)

Date 08-04-2003 19:58

Hi MikeR

Your weld composition will end up with around 0.5% Si and 0.4% Mg. This is a very crack sensitive composition, meaning that you will struggle to weld this without cracking, especialy in a constrained weld.

I believe that you will have to add enough of a filler such as 4043 or preferably 4047. With 4047, you will need to add less filler to get out of the crack sensitive zone.

Hope this helps.

Regards
Niekie Jooste
Fabristruct Solutions

By MikeR

Date 08-04-2003 22:12

Thanks. Oh well, filler it is.