Reply With QuoteAnd a pic if that helps. The electrode above the pipe which is laying on the plate and a joint (if possible) formed where the dot is. 1.jpg
Here we go, post number one. I was wondering if anyone had any experience of holding a copper pipe against a flat steel surface. I have a few ideas such as spot welded clips, double sided thermally conductive tape between the two and maybe even spot / resistance weld the pipe directly. Spot welding the pipe directly appeals but I'm not sure if it's possible. The thought is to attach a ground clamp somewhere on the steel plate then place an electrode on the pipe edge in the hope that the current takes the shortest route directly underneath to where the pipe touches the plate to form a joint. If anyone has come across this before I would be keen to hear how you resolved it. Thanks.
And a pic if that helps. The electrode above the pipe which is laying on the plate and a joint (if possible) formed where the dot is. 1.jpg
Is there a reason why you can't put a bracket around the copper pipe? It' looks like it's not covering a very good area for a stable joint.
Seam weld via electrical resistance? How about one of those over-center toggle clamps - electrically isolated of course.
What is the pupose of placing the copper pipe between the electrode and the steel pipe? From your picture and description it appears you are simply trying to transfer a voltage and/or current from the electrode to the steel plate. Why not simply place the electrode in direct contact with the steel plate? The steel plate is going to provide the primary resistance to the current flow so the lower resistance of the intervening copper pipe in your picture will not have any beneficial effect on current rate. If the pipe is spot welded to the plate then the spot weld potentially presents the highest resistance point in the electrode to plate circuit with a resulting voltage and current drop.
I should have been more descriptive, apologies. The basic intention is to try and cool a surface.
To do this a refrigerant is pumped through the copper pipe and takes heat of the stainless steel surface. The pipe 'zig-zags' across the surface. A good bond between the pipe and the plate is essential for best thermal transfer. There are options such a thermal compound between the pipe and plate, using clips or a combination of the two. However, in an effort to eliminate a wet process (thermal compound) and reduce processing time I did hope to consider joining the pipe directly to the plate by resistance welding at intervals along its length.My concerns are the dissimilar metals may lead to damage on the copper pipe enough to loose it's pressure integrity.
The clipping down the pipe is worth following, I could spot weld the clips down over the pipe onto the surface. I would prefer to do this robotically to ensure speed and repeatability, if possible a automatic pipe clip loading at the head would be nice. All of the pipe clips I’ve seen so far such as ‘P’ clips have a hole for a fixing.
I’m aware that quite often inspiration can come from others having done or seen something similar elsewhere so I thought I would ask the community here for help. I appreciate your replies so far, thanks.clip.jpg
Last edited by Siftersam; 08-29-2013 at 09:55 AM.
OK, now I understand the base issue and that the diagram displayed the welding process. First, with regard to the welding resistance welding procedure I am concerned that due to the much higher conductivity of the pipe, particularly in the case of SST, which has a low thermal conductivity, the applied current will be evenly distributed along the entire length of the pipe.
Now as for the base problem of cooling the SST plate with the copper pipe with or without any intervening thermal transfer agent. This is where I have some major concerns about the base concept's success. First let reveal that I spent the last 15+ years of my engineering career developing ASME Section VIII pilot operated pressure relief valves for superheated steam services up to 1000F (538C) and had to deal with the issue of providing reliable even heating from the main valve body mounting to the attached pilot body assembly and have performed significant testing in this respect. In my applications the problem was achieving an acceptable surface contact to provide a reliable heat transfer between two mating machined flat surfaces and we found this was impossible to achieve with even well finished flat surfaces without an intervening transfer agent (In our case the best solution was achieved used a soft graphite sheet material). Based upon this background, my major concern is the very small actual heat transfer contact area inherent in placing a round pipe on a flat plate. Due to the high heat conductivity of the copper pipe vs the low conductivity of the SST plate I am concerned that without very close spacing of you pipe runs you will end up with narrow cooled strips under and adjacent to the pipe contact lines on the plate with unacceptably high temperature areas between the pipe runs.
With all of the above concerns now iterated, the only suggestion I might offer that could improve the cooling heat transfer between the plate and pipe would be the use of flanged horseshoe shaped long strips of SST (think of a series very long horseshoe shaped pipe clamp the length of each of your straight sections of pipe) that could then be spot or resistance welded at several point along their lengths to tightly grip the cooling pipe to the plate and provide a larger surface contact for heat transfer on both the pipe and the plate surfaces; but, even with that arrangement I am concerned about good heat transfer from the plate to the pipe. It just depends on how much and how reliable cooling you need for your application.
OK, now I understand the base issue and that the diagram displayed the welding process. First, with regard to the welding resistance welding procedure I am concerned that due to the much higher conductivity of the pipe, particularly in the case of SST, which has a low thermal conductivity, the applied current will be evenly distributed along the entire length of the pipe.
Now as for the base problem of cooling the SST plate with the copper pipe with or without any intervening thermal transfer agent. This is where I have some major concerns about the base concept's success. First let reveal that I spent the last 15+ years of my engineering career developing ASME Section VIII pilot operated pressure relief valves for superheated steam services up to 1000F (538C) and had to deal with the issue of providing reliable even heating from the main valve body mounting to the attached pilot body assembly and have performed significant testing in this respect. In my applications the problem was achieving an acceptable surface contact to provide a reliable heat transfer between two mating machined flat surfaces and we found this was impossible to achieve with even well finished flat surfaces without an intervening transfer agent (In our case the best solution was achieved used a soft graphite sheet material). Based upon this background, my major concern is the very small actual heat transfer contact area inherent in placing a round pipe on a flat plate. Due to the high heat conductivity of the copper pipe vs the low conductivity of the SST plate I am concerned that without very close spacing of you pipe runs you will end up with narrow cooled strips under and adjacent to the pipe contact lines on the plate with unacceptably high temperature areas between the pipe runs.
With all of the above concerns now iterated, the only suggestion I might offer that could improve the cooling heat transfer between the plate and pipe would be the use of flanged horseshoe shaped long strips of SST (think of a series very long horseshoe shaped pipe clamp the length of each of your straight sections of pipe) that could then be spot or resistance welded at several point along their lengths to tightly grip the cooling pipe to the plate and provide a larger surface contact for heat transfer on both the pipe and the plate surfaces; but, even with that arrangement I am concerned about good heat transfer from the plate to the pipe. It just depends on how much and how reliable cooling you need for your application.
I’ll be running some trials soon on various methods:
· Attempting to resistance weld the pipe directlyto the plate. Don’t hold out much hope for this as I believe it will damage thepipe and significantly reduce the pressure integrity.
· Pipe clips at intervals without any form ofthermal transfer compound between pipe and plate.
· As above but with a thermal transfer compound.
· Lengths of strip along the pipe length in twosections. V section and ‘horseshoe’ section.
Thanks for any advice here.
One additional late caution regarding the welding of the pipe directly to the plate. Don't forget to consider the possiblity of fatigue stress failure of the welded connections due to the differential between the thermal expansion coefficients SST and copper, particularly if there will be heating and cooling cycles.
What about a brazing operation to create a joint between the two parts? Like JAlberts mentioned, you do have a very small area of contact between the tube and plate so your heat transfer will be limited.
Kyle
Use square or rectangular tubing. It'll give you more contact to your plate. I'd braze or silver solder it to the plate. If you need round for connections then braze a fitting to the end. We make coils for induction heating here now and then and use square tubing for these. Gives a better field... We poke a punch in the end of the square tube and stick a stub of round tube to make brazing a fitting on the end easier.
Might take some trial and error getting the process down. If the steel plate is pretty thick you might want to soak it with heat before you even introduce the copper. If you're doing zillions of these I'd look into induction heat with tape type solder or brazing material and a set of rolls and a quenching arrangement. Feed the plate past an induction coil while a square copper tube with the solder tape attached is being laid and held firm to the plate via rollers for a long enough period of time for the solder to melt then pass through a spray of water. Follow that?
Good Luck...
Bob
Nice idea but the 'non pipe' side of the stainless plate must be clean. Brazing will discolour the surface and require a second op to rework it. Resistance welding a plain copper stud to the surface works so far. A clip is then passed over the stud and retained with the type of washer that grips in the bore. This is ok but as the operation is repeated around 100 times per unit I would like to look into resistance welding the clip (or pipe) directly onto the surface and maybe automating the operation with a robot and some form of clip feeder.Originally Posted by kjoiner
Gut feeling even from the start tells me joining the pipe directly is probably a no go'er. As JAlberts mentions, there are potential problems with thermal expansions, reliability, initial damage to the pipe etc. However, I thought it worth asking the question.
Posting Permissions