??? How to Make Long Welds Without Expansion Cracking ???

I would preheat he whole piece to around 500 degF, and then run the weld either pulsed MIG or pulsed TIG. I have run 4 foot seams like that in one shot using pulsed TIG. Pulsed MIG would be faster. If you are using standard MIG or TIG, just skip stitch it to break up the heat.

Minimizing heat absorption isn't the goal. The goal is to reduce thermal differentiation. You want as even a heat as possible so you don't have really hot areas and really cold areas. Preheating helps a lot as does skip stitching.

Would you use a O/A torch rosebud for this? The overall work piece is 16' long so putting it in the oven would not be very practical.

How critical is the heat range. You say about 500, but aluminum cools very quickly. Would you maybe heat to 600 and try and get the entire work piece done as quick as possible before the temp dropped under 500, or keep stopping to reheat? Also, would a laser temp gage be worth while investment for this type of work?

Standard MIG. Preheat and stitch then. I assume you mean weld an inch and then skip a couple inches and weld another inch?

Thanks Ernie. As always you are a great help.

In class, a guy brought in some 1/4" aluminum and experimented with trying to stick weld at T with it. Without preheating, it didn't really work at all - mostly just a lot of cold lapping happening and the weld just broke off with almost no effort.

Next time he tried pre-heating. I don't remember the numbers, but I think one time he tried it to about 400 degrees and the next time to 200 degrees. One worked fine, the other produced a large crack along the back side of the weld at the bottom of the material. So, in that case, I would say heat range was fairly critical.

I don't remember exactly what he did, but he might have had the T sitting flat on a metal welding table and that might have caused a lot of heat to be sucked out the bottom which might have caused the crack???

My point is that preheating can be very important for aluminum because of the fact it expands and contracts so much with heat changes and the heat range can likewise be important for a the application.

I've heard that it can be important to insulate the work at times to keep it from cooling. Not sure what would make sense for your application but I think there are blankets that can be used to control the cooling.

I understand that make electric heating blankets as well for pre-heating but I don't know anything about them (cost, size, etc).

For the type of weld you are doing trying to weld long strips of thin aluminum to a frame sounds like a disaster waiting to happen. I have no clue how they normally build things like those big aluminum boats and deal with the warping and cracking problems but it sounds like it could be a serious problem.

I own one just so I could get a better feel for temperature ranges in the work I was doing. But I find it hard resisting buying any tool if I can even half justify owning it. :)

They guy experimenting with the aluminum was using one to measure how hot it was getting for preheating and to see how fast it was cooling off. I would say it would be a wise investment if you want to continue to do the type of stuff you are talking about here.

The cheaper ones don't have as large of a heat range. You have to spend more money to get ones that go to higher temperatures. I think I spent around $100 for one to go up to around 1000 deg.

I've found it to be useful in other applications as well such as testing the temp of the fridge and hot water heaters etc.

You kinda sound like a buddy of mine. He wants a thermal imaging camera (about 7 grand) so he can check for energy loss locations in the rooms in his hotel. Fire up the airconditioning and let her rip for a while then scan the room with the camera and find all the hot spots. LOL.

HF has one that claims a range of -4°F/-20°C to 968°F/520°C on sale right now for 29.99. I just happen to see it in the store the other day for a lot more. I don't know if it would hold up very well, but you never can tell. It might give me some idea of where I am at for temp anyway. Should be enough since 5052 melts at around 1200°F anyway. Some HF stuff seems to hold up and others fail before you get it out of the package. They have a good return polciy though.

I have this one bought it for powdercoating. Works great so far (about 6 months old). Accurate to about +/- 3F as far as I've found.

--

-Smitty Somerset, PA USA

Thanks Bill. Good to get some feedback. I just ordered one to try.

I have the Fluke 62 mini which sells for about $90 and a temp range to 500C so it's basically in the same class as the harbor freight unit. The accuracy and range seem to be about the same except the Fluke seems to go to slightly lower temps (-20F vs -4F).

Here are the specs:

(FlukeProducts)

The fluke also has a max temp feature so as you scan an area, the max temp seen will be saved and displayed at the same time it's showing the current temp. But other than that, the units look basically the same and the HF unit is less than half the price.

Another thing I've used it for that you might be interested in is checking the temp of my oven when powder coating.

In reading about these units to make this post, I've learned something new.

The shinier the material, the less accurate the reading. Measuring some materials like aluminum, can be tricky because of this. I've seen some odd numbers with my unit in the past for materiel like aluminum but didn't understand what was going on. After some experimenting with the over this morning, I think I have a better understand of the problem.

There's a parameter of material called the emissivity which is a measure of how much radiation it will give off relative to a perfect black body. It's a number less than 1. A value of 1 is a perfect ideal back body which absorbs 100% of the light that hits it (and looks black because of that).

Most typical non-shinny materials have an emissivity over .90. Here is a list of some values.

If the material you are trying to measure has an emissivity near 1, these IR temperature units will give a fairly accurate reading. But for values much less than 1, the accuracy will not be very accurate. The more expensive Fluke units allow you to enter the Emissivity of the material you are trying to measure so you can compensate for it. The cheaper units like Mine and the one from HF don't adjust. The HF manual says that to measure the temperature of shiny object, put duct tape on it and measure the temperature of the duct tape instead (one more use for duct tape!). The Fluke manual suggest you paint it black. :)

But none of the manuals talked about how much error to expect, or what the error would be like. So I did some reading to try and understand the issue and I did some testing with my unit.

Walking around with my unit measuring the temp of dark objects and shinny metal objects, I couldn't see any real errors. The shinny objects were within a degree or two of all the other objects around them.

So I put a few objects in the oven and heated it up. I put in an alumuminum cookie tray, a piece of very shinny aluminum foil, and a cook tray from teh toaster over which seems to be bare aluminum on one side, but blackish non-stick coating on the other side. Heating the over to around

350, I opened the door and measured the temp of these items as there were sitting inside the hot oven. Again, no real error noticed. There was temperature differences of about 30 degrees depending on what part of the oven I was pointing or what item I was pointing at but that seemed mostly normal temperature variations - and because the door was open, the temps were dropping fairly quickly at the same time. Nothing really indicated to me that the shinny aluminum was creating any big error. With the exception however that as I scanned over the shinny aluminum foil, the max reading of my IR thermometer would sometimes catch an oddly very high reading (like 500+ when everything else was in the 340 range).

The emissivity of Aluminum can be as low as .04 from the table on the above web page, and less shinny aluminum is in the .1 range. Which means that the amount of radiation aluminum gives off due to it's heat is 99% less than a black body should. If the IR thermometer is measuring this radiation, and calculating the temp based on what it reads, then the indicated temperature it seems to me should have been 99% below the real reading (relative to absolute zero!!!). So why wasn't I seeing a huge error????

After more reading, I finally grasped what was going on.

The more reflective a surface is, the more it will reflect the radiation from _surrounding_ material, but the less it will radiate due to it's own internal heat. How much it radiates due to it's own heat, is exactly inverse to how much it reflects. So material with a .1 emissivity coefficient will radiate 1% based on it's own heat, and 99% based on the the heat of the radiation that is falling on it from surrounding objects!

So here's what is happening as I now understand it. When you measure material with a high emissivity, you are mostly measuring the heat of that object. That is, the IR radiation it is giving off, is mostly due to it's own surface heat. But when you measure a material with .5 emissivity, half the radiation you measure is based on the internal heat of the object, and half is just reflected radiation from the objects around it. So if the object you measure is roughly the same heat as all the objects around it, you will notice no error at all.

This is why when I measured the objects while they were in the oven, the temperature looked all about the same. And it's why when I walked around the room, where everything in the room was about the same temp, the shiny stuff and the non-shiny stuff all measured about the same.

So, to test this, I take the oven tray which is has the dark coated non-stick coating on one side, and bare metal on the other, out of the oven (holding it in a oven mitt) and measure the temp of both sides.

The dark side measures 327 F (basically correct for something just coming out of an oven set to 350 F), but the shinny side measures 145 F! This was in a room where the temp was around 75 F. The temp was dropping as I measured it, but as I flipped it back and forth and checked the temps, the dark side was consistently high where as the shinny side was consistently far lower.

A little math based on the temp diff shows the shinny side of that tray had an emissivity of about .3.

This explains why at times I've measured the temperature of a hot piece of aluminum and got numbers far lower than expected (I think I burned myself once when the IR therm showed something in the 150 range so I thought it would not be too bad to touch - but it was probably more like the 300+ range and wasn't so nice to touch).

So, the whole trick is that when you try to measure something shinny (like aluminum - even dull dirty oxidized aluminum like my toaster oven tray), you end up measuring more of the reflection of the stuff around it, instead of the temp of the actual unit. So if it's surrounded by material of the same temp (like in the oven), you get a fairly accurate number, but if it's temp is very different than what's being reflected from around it, like when trying to measure preheat temp of a piece of aluminum sitting outside in the open air, there's going to be a lot of error.

I guess as long as you are aware of how the error works, and can approximate an emissivity value for the material, you can just adjust based on the temperature of the stuff around you. So, for really shinny material with an emissivity of .1 (clean aluminum sheet) the difference from room temp to the temp indicated will only be 1/10 what the difference should be. So if it measures 10 degrees above room temp, the real temp will be more like 100 degrees above room temp.

Some experimenting by painting the surface flat black (or using duct tape) and measuring the temp of the flat black part (the real temp) vs the rest will be an easy way to figure out the error for any material.

I've been thinking. The outside edges of the 1/8" sheets is trapped and welded in place. I am wondering its the simple cooling and contracting that is causing the cracking as much as temperature differential? Here is my idea. Cut a nice chamfer on the facing edges of both the 3/8 and the 1/8 so that they come together in a wider V. (I can do this with my die grinder fairly easily.) I think that would guarantee a better amount of fill and weld directyl between the pieces. The V I have now is minimal. Then drive some small pins in between the 3/8 and the 1/8. Maybe something the size of some welding wire. This will put the 1/8 under a constant state of pressure. Preheat, stitch, pull pins, and finish. Hopefully this way there will always be a slight pressure into the weld, and there will never be tension on it.

Ah, I just realized something else.

Without knowing the physics, I was thinking the IR radiation from objects to be be fairly constant based on their heat. So if there was an issue with shinny material reflecting IR light, that would only add to the amount of radiation coming from the material and cause an IR thermometer to register too high. But then reading about it, I learn that different objects give off different amounts of radiation directly linked with how much they reflect. So the more IR radiation an object will tend to reflect, the less it will also tend to radiate from its internal heat by the same factor. So aluminum, which has a very low emissivity value, reflects a lot of IR, but also, doesn't give off much when it's heated. And duh, that's what just clicked with me. Unlike iron, which glows very nicely when it's heated, aluminum doesn't glow when you heat it (which makes it that much harder to figure out when you are about to melt a big hole in it).

But, according to what I've read today, it should glow with basically the same frequency spectrum of iron, but with just far less intensity (that is, adjusted by it's emissivity value). I don't think I've ever noticed any glow from aluminum when I've welded it. Is this just because I was always in too bright of a light when heating aluminum to notice any glow from it???

Now I feel the need to go do a science experiment and heat up some aluminum and turn off the lights! :)

Hmmm.... I wonder if the black soot from heating it up with a rosebud torch would fix that problem?

I bet it would. And it's such an easy solution that works well for high temp. You will have to do some testing when you get your HF unit and let me know how the temp of the soot compares to an area without soot. (I don't have an O/A torch to test with).

Its a shame you don't live down the street. I actually have spare O/A bottles, and everything else spare except a spare cutting head. One of these days I need to test it all and see what is good and what isn't. I think I have three sets of hoses atleast, and some various gages.

Oxy/Propane is the way to go. Just hook up your acetylene regulator to a BBQ propane bottle. Propane makes rosebud tips really happy.

Wool army blankets work well to keep in heat. So heat a large area until a 600 degreeF tempilstik melts all over it, then throw blankets over as much as possible and start welding.

If the aluminum drops to below 400 degrees, reheat and continue.

Tempilstiks are very useful for stuff like this.

Divide an area of the seam into even sections about 3 inches long. If these were labeled 1,2,3,4,5... and so on you would follow a weld sequence of 1,3,5,7,..., and then go back and weld 2,4,6,8.... This is skip-stitching.

Backstep skip-stitching is a little more complicated where you go 2 to

1, then 4 to 3, then 6 to 5... Then go back and weld 3 to 2, then 5 to 4, then 7 to 6...

I am just not picturing that.

>

Here is a diagram:

What about Tempil sticks? Or do those only work on steel?

-- As Iron Sharpens Iron, So One Man Sharpens Another. Proverbs 27:17

That's what Ernie suggested. Tempil sticks.

I've never welded aluminum (not a qual I've been sent in for yet), but I do know that it oxidizes fairly quickly. I think our standard is to clean the joint components, then weld it up within 8 hours. If the welding takes longer than 8 hours, we gotta stop and clean the oxidation off again (I'm not positive, but I think it's 8 hours- I know it's no more than 16 hours).

Question 1: Will pre-heating cause the aluminum to oxidize faster?

Question 2: More weld is not necessarily better- if you increase the included angle from 90 deg to, say, 120 deg, don't you increase the weld shrinkage and likelyhood of cracking?

Question 3: If you have enough overlap of the 1/8th sheet over the frame, could you cut holes in the sheet and plug weld sheet to frame, then start welding up the edges? Would this distribute stresses better?

Last one: How hot would a propane torpedo heater get the sheet? If you could heat it on one side while welding on the other- could this work?