Strength of Aluminium Weldments

[britishrocket]

Hello All,

Here is a thing. There are a fair few TIG welders amongst us here, and we use these skills to produce devices that generate thrust, be these turbines or rockets. So, we want low weight to make the most of the thrust we can generate. That is inevitably going to lead us down the road of using aluminium in our builds.

I'm currently contemplating making a combustion device and I've been looking at 316 stainless steel as the primary material. I cannot, however, get away from the idea of using aluminium. It is so much easier to machine, and has great weldability. It also has a very high thermal conductivity, a useful property in many combustion devices. On the other hand, 316 welds well but the weld can be expected to be at least as strong as the parent metal, all other things being equal.

I am sure everyone here is familiar with 6061 T6 in it's various forms. In the United Kingdom 6082 T6 is more well known and popular. In the T6 temper both these alloys have a UTS in the region of 300MPa.

Once welded though, the heat affected zone (HAZ) is annealed and the strength goes down to that of the T0 temper, typically an 80% drop. Now, we could design this into our device, but that would tend to increase bulk and therefore weight, somewhat detracting from using the alloys in the first place.

Reading around this subject, I have seen that there is a certain amount of age hardening that can take place that takes the HAZ up to the T4 temper, so not such a huge deterioration. How much time seems to be a matter for debate, days to weeks are quoted. There'd be no way to be sure. A better bet seems to be precipitation hardening (I think) in which the whole welded piece is heated to 200 Celsius or thereabouts, then cooled slowly. This is supposed to return the HAZ to T4 and seems a lot less hit and miss.

So I was wondering if anyone here has any experience of non heat treated welds failing? Or if they have heat treated a weld they have produced, and seen a strength increase, evidenced by testing?

I'd be interested in any thoughts on this. I would like to do a few trials myself, weld up some butt joints and heat treat some, then rig up a makeshift tensile tester using an enerpac jack.

Thanks,

Carl.

[mitch]

I haven't had any of my mig welds fail (non are heat treated). I use mild steel though, and whenever I have to weld something structural or load bearing, I make sure to double pass or do a good job lol. That seems like an interesting experiment though. I am considering buying a tig sometime soon here. Post some results if you go through with the tensile testing!

[britishrocket]

Hello Mitch,

Thanks for the comments. The thing with aluminium that is completely different from steel is that due to it's metallurgy, the area surronding the weld is substantially weakened. There is no way to avoid this. I'm not sure if you are in the UK or US or where you are, but say you welded a piece of 6082 T6 (I think 6061 is more common in the US) The heat affected zone of the weld would not have a UTS of 300MPa anymore, like the rest of the piece. It would instead be something in the region of 80% less. That is to say, it will have gone from the T6 condition to the T0.

I have read that it can "age harden" back to T4 after a period of time , but there is significant disagreement of how long. The heat treatment can take it back up to T4. You can get it back to T6 but this involves getting it very hot, say 500 Celsius, then quenching it in water, then a slow heat to 200 and a slow cool. I don't have an oven that can get up to 500 C let alone one big enough to get a good size part in. I also think that quenching a complex ally fab from 500 C in water could cause distortion....

So it seems the best way would be to get it back to T4 temper by heating to 200 C (do-able) and having the part designed with the T4 strength properties in mind.

So in a nutshell, that is why I want to do the tensile tests, to see if the heating to 200 C really works. And I will definitely post the results of any tests here! At the minute my plan is to continue my rocket combustion chamber in 316 stainless. I'd like to use welded aluminium bits in an engine design but I really want to make sure that the welds would hold up to the pressure.

[gtbph]

Hi Carl,

I would suggest using EN AW-5083 for example.
It is an alloy that is not hardenable by heat treatment, its strength of around 300MPa comes from the chemical composition alone, so it doesn't loose strength when welded.

In a rocket engine I think the alloy will be exposed to some "heat treatment" automatically, at least to a certain depth. I don't know if this would harden or soften a heat-treatable alloy, but with a non-heat-treatable alloy you would know that not much will happen.

Regards, Alain

[britishrocket]

Hello Alain,

Thanks for the comments! Very interesting stuff this 5083! According to the Aalco datasheet, it has a UTS of around 300MPa even in the 0 temper condition. The drawback that I found was that it is not recommended for service above 65 Celsius. I'm not sure why this is.

My engine design is to be regeneratively cooled. I have put my heat transfer calcs on my build thread in the rocket section here and on my blog. For 316 stainless, which has a thermal conductivity in the temperature range of interest of about 16 W/m^2 K, I can keep the temperature down to about 200 ish Celsius. Fine for stainless and probably OK for T4 6082 aluminium too. If T0 6082 was taken to this temperature then as you say it would do the heat treatment, once it had cooled, but the parts would have to be designed to take the operating loads in the T0 condition, which is how they'd start out. So sort of negates any "automatic" heat treatment that might come about.

Obviously aluminium has a much higher thermal conductivity than stainless steel. I have not plugged the value in to my heat transfer calcs to see what temperature I could get the thing down to, but I'm sceptical about achieving more than a few more tens of degrees. Getting it to 65 Celsius or less seems impractical. Plus once the engine has run and the fuel is no longer circulating as a coolant you get hot soak back, which with 316 there is a decent margin for this to be acceptable, whereas with an alloy that can't go above 65 Celsius this might be a problem. I actually have a plan to cope with heat soak back on the test stand, which is to by pass the injector and have secondary cooling water flow through the cooling jacket after the burn is finished. still doubt 65 Celsius is possible though.

Thanks again Alain and Mitch for your comments. Knowledgeable people and thought provoking postings are exactly the reason I use this forum!

Carl.

[gtbph]

Hi Carl,

65°C is not much, but I found other data saying long time exposure up to 120°C is ok, and short time exposure up to 180°C. For 6082 the same site gives 120°C / 160°C.
The site is in german unfortunately: 5083 6082

If you have not ruled out aluminium yet, I found something in a pdf about heat treating 6061 that could be of use for you:


This is about the effect of temperature on 5052:


The full pdf is here.

Alain

[britishrocket]

Hi Alain,

Thanks for that, the paper was very interesting. It mentioned 2219, which is a grade of aluminium whose strength is not affected by heat at all and can be welded. I should have known this really as I have an aerospace background. The 2xxx series alloys are the Rolls Royce "RR" series, developed in the 20s and 30s for aero engines and ultimately used as the primary structural material for Concorde. I know them best by the name "Hiduminium". This was from the company that RR set up to develop them, called High Duty Alloys. Hence high duty aluminium became hiduminium.

After that brief history lecture it is back down to earth, because 2219 is extremely hard to get hold of in tube form, it is mainly produced as sheet. Not surprising really as the 6xxx series are the extrudable ones. So it is back to good old 6082 again.

If - and it is a big if - it is possible to heat treat the welded part to the T4 temper, it should achieve a UTS of about 205MPa. According to MIL-HDBK-5J, 6082 T4 derates to 90% of UTS at about 100 Celsius. This 90% value is 184.5 MPa. This sets the bar as this value is the minimum at which the combustion device wall dimension can stay the same as the 316L version, namely 2mm.

At the moment the cooling calcs I've done with 316L as the material give a hot gas side temperature of 300 celsius and a coolant side temperature of about 240 celsius. This will have to come down some, but it has to be borne in mind that the thermal conductivity of 316L is 16 W/m^2 K, whereas that for 6082 is ten times more.

So I'm going to plug some numbers into my cooling calcs and see what happens, it may yet still be do-able.

Carl.

[britishrocket]

Hello All,

Well, it looks as though I am on a hiding to nothing with regard to welding an aluminium tube chamber. I had hoped that if the component could be heat treated then the whole of it would assume T4 mechanical properties.

This isn't the case though, I was forgetting about the weld metal that is of course an alloy of say 6082 and 4043. A friend who has experience in designing and making aluminium weldments told me that they allow 101 MPa for the weld itself. So derating this for the temperatures involved in a rocket chamber makes it a non starter.

Carl.

[britishrocket]

I've been doing a lot more research on this topic and I have just put up a sizable post on my blog:-

www.britishreactionresearch.blogspot.co.uk