Welding Tips and Tricks

All

Hello to everyone here on this forum, this is my first post so hopefully you can help me out.

I have a current project on the go where I need to weld a aluminium flange around 10" dia by 1" wall thickness probally 6082 to a cast auto gearbox casing. I am unsure of the casting material, but I would guess it is a high silicon casting alloy similar to LM25.

Firstly, what pre-heat should I aim for, 300-350°C and above?
What cleaning should I do or what surface prep would be best and would the pre-heat compromise any cleaning already done on the cold material?
I intend to use a 4043 5% silicon consumable 3.2mm dia but would anyone recommend anything else, maybe high Si content?
I will use a Zirconiated electrode with gas lens and have the option of an inverter set or an older MUREX square wave. Any preferences?
Finally, do you think any a post weld heat treatment will be needed?

Thanks in advance

look through the tig forum, the cast aluminum surface prep has been covered a few times recently.

the 1" thick flange is going to require a lot of heat, and most likely multiple passes. you can try a helium or helium mix, but unless you have a 400A machine, you're going to have to bevel pretty deep and do a multi-pass weld.


are you adapting a tranny to a different engine?

if this is going to be in a racing application, you might contact a place that does shot peening. you should be able to get some additional strength out of the weld area by having it peened. most aluminum isn't heat-treatable (I don't know about cast, I haven't worked with much of it) but work hardens well. you should be fine as welded, but if this is for a drag car or autox, I'd probably do some looking into my options for post-treatment.

Thanks ogorir,

The trans is the original for the engine and it is a longitudinal design i.e. for a rear wheel drive car. I intend to convert this trans to a transaxle as the project is for a mid engined car. The original trans is to be modified to accept a set of gears at its rear which turns the output shaft through 180° degrees and sends a transmission shaft back up the outside of the casing into a billet differential housing that I have machined. It is this differential housing that needs to be welded to the original bell-housing. The differential then puts two drive shafts perpendicular to the original input shaft.

I have managed to reduce the mass of the diff housing to a minimum without compromising its integrity by using Finite Element Analysis ensuring the stress levels and distortion levels are within acceptable parameters. By reducing the mass I am hoping the heat required is lower and it will be easier to maintain the original housing and billet diff housing at an acceptable pre-heat temperature.

I do have access to a big AC machine and hopefully a much better welder than I am. All I need to do is try and find out what the best procedure will be, for pre-heat temperature, consumable specification, gas consistency (argon, helium or mix). I am sure if I get this right I have a fighting chance of bring this off.

The final product will not be used in anger, it will be simply used to prove the design of the transaxle ensuring the thing actually works! I will then strip the transaxle and use the welded part to produce a pattern then have a part cast.

I have heard that I need a good pre-heat as if I try to heat the part too quickly with the torch I will pull out any contamination from the casting which will then in turn make the weld crap! The cast bell housing looks to be investment cast, so I am hoping it will be a good quality and not a dirty casting you sometimes get from sand casting.

Anyway, thanks for the info so far and if anyone else has been down this route and has any advice then please let me know! If not i will ensure I document what I find and let everyone no, even if it all ends up on the workshop floor

Cheers

Steve

Is it possible to use studs and bolt it together?

Hello sschefer

No it isn't possible to bolt together, which is unfortunate as I am a big believer in engineered components that fit together like a rifle bolt. Maybe not the best quote on a welding forum

Anyway, here is a bit of an update that I did today.

I machined away part of the casting and had a go at welding the offcuts. These were bits of the casing about 5mm thick of various shapes but on average about 50mm x 50mm.

The surface was cleaned by removing the skin and then cleaned with some cleaner I usually use on wrought. The material wasn't pre-heated for the first trial. The machine I used was an inverter set, AC selected, no pulse and 100Hz frequency. I set the current initially at 115amps and would control the current on the pedal. I had a 2.4mm Zirconiated tungsten, ground to a point, pure shield argon and about 12-15 cfm flow rate. I also had a 2.4mm 4340 5% silicon consumable.

The initial strike up looked promising and I started adding rod immediately, however within 20mm of travel it all went wrong! The puddle skinned over and the turned into Aero bar, full of bubbles/craters. The end of the tungsten balled up real funny and not on the end of the point but slightly up the side which was a bit odd. I re-ground the tungsten and had another go. This again was unsuccessful and I couldn't even get a puddle started as it just skinned over straight away. I changed the tungsten size to a 3.2mm and had another go, but no joy. Tried to add a bit of pre-heat on the material, probably up-to 150-200°C but again no joy, tried changing the frequency to 200Hz, great buzz but no real change.

I then had a bit of a play with the pulse, but it didn't make enough change to warrant any perseverance.

I then wound the current down to 85amps and again controlling it on the pedal, I very lightly started an arc. This made the most difference and I was able to get a very small puddle or at least a bit of surface wetting to get the rod to flow into it. However I am not sure this is right as the surface sometimes skinned and sometimes wetted so it was really inconsistent. Not great!

What was noticeable though, was when the material was pre-heated slightly and if it was heated too fast, the material seemed to erupt in small locations as if something was trying to get out So, I am guessing this was why the hot arc was causing the material to bubble and skin and create Aero bar!

So no real success today, and would like to know if any guru's on here can give me any pointers!

that bubbling and skinning is normal. you're likely going to have to weld, grind, weld, grind, weld a lot. castings are porous and absorb oils and can oxidize further from the surface than forged or billet parts.