Unclesyd,
I read your post very carefully, as I do all your posts. I have a great deal of respect for your experience, the quality of your posts, and the specific successes you mention here.
However, since this forum is a learning place for all the readers, I am going to present a contrarian point of view, and the readers can then take what is presented, and make up their own minds as to which way they might want to go.
I, too, like you, over the last 35 years, have participated many times in the (roll first, then weld, or weld first, then roll) argument, and have heard it presented both ways quite elequently by very knowledgable people at that.
Then, only a few years ago, while in the shop of a major heat exchanger and condenser manufacturer, inspecting a titanium tube/tubesheet condenser for a client, upon noticing that they were welding after rolling, I asked the quality control manager why, as I knew for a fact that they used the opposite procedure for their high pressure Hx's.
He gave me the explaination regarding the blowout of the gas pocket that I mentioned above. Manufacturers are always concerned with shop production rates, and cannot afford to take the time to have to go back and rework tube end seal welds in a HP Hx containing literally several thousand welded tube ends. So, they were using the method that they knew would get it right the first time.
These were Hx's that can exceed 6 ft. in dia., and have tubesheets as thick as 27 in., and can be over 40 ft in length, and weigh enough that they have to be shipped on specialty equipment.
Moreover, since reading your post, I have also pulled a couple of my clients HP FWH specs. These are major Electric Utilities, where the specs basically originated with AE's like Stone and Webster, and/or Sargent and Lundy, to mention a couple, and find that these specs all call for welding first, then (1) visually inspecting, followed by (2) an air test (35 psig was the requirement on one), followed by (3) a dye penetrant test.
If all that passed OK, then, and only then, the tubes were to be expanded, and the tube seal welds were dye penetrate tested again after the expansion step. Once the shell was pulled on and welded, then the tube/tubesheet welds were to be tested yet once again, with nitrogen or air.
My experiences with several of the major FWH manufacturers, some of whom are now gone, has basically been the same. They all weld first, and then roll, except, of course, with Titanium. Senior Engineering, one you mention, now called Thermal Engineering Inc., TEI is but one of them.
In your case, you may have had excellent welders, or welding procedures that anticipated the gas pocket blow out problem, and accomodated it, or compensated for it in some way.
Or, it could just be that the expansion roll itself provided enough sealing that any flaw in the tube end seal welds was not detected. After all, many more Hx's are built 'rolled only' than are built welded and rolled (or vice versa as in your case). Whatever, it worked for your plant. And, additionally, explosive expansion normally expands most of the full length of the tube in the tubesheet, as opposed to the 2-1/2 in. manual roll that was done years ago, which helps with the sealing of the tube.
Another argument often heard is whether the weld is for sealing only, or is to contribute to the strength of the joint to resist tube pull out.
I have an anecdote regarding that. A major FWH retubing outfit once went into a plant to retube a huge HP FWH on a supercritical boiler which could not be removed from its location on the mezzanine level due to a lack of crane capacity in that area of the plant. They found one complete row of tubes that had been seal welded, but had never been expanded. (and they were having beau coup trouble with removing the tubes that had been expanded into the 27" thick TS). This row of tubes had survived all the years it took the heater to die a natural death, and were not part of the failure mechanism that occasioned the need for retubing. So, the seal weld in and of itself is quite strong as well, but in reality it is the expansion that is designed keep the tube in the tubesheet.
Speaking of tube expansion, the major FWH manufacturers, if given their choice, (sometimes customer specs govern) will use explosive expansion. It is much faster than the other methods we have discussed, given the ability to shoot whole rows or combinations of rows at once. It is very consistent, notwithstanding the negative comments on the website we discussed, and produces good results.
I have one other thing to throw out, that is my opinion and only my opinion, although others share it with me. I want the readers to be aware of the potential problem.
For HP Hx's, with thicker tubesheets, and with heavier wall tubing of the harder tube materials, (Stainless steels, carbon steels, etc.,) I don't recommend tube hole grooving. Let me state categorically that with softer tube metallurgies, brass alloys, Cu/Ni alloys, etc., grooving is mandatory!!!!
The consideration I want to present is the potential for cracking in the tube after expansion due to the sharp corner (bevelled or not) of the groove. With full length expansion by whatever method, the grooving is not necessary to obtain the full tube pullout strength.
Unclesyd, I do, however, like your 3 groove approach. I do agree with the concept that if two are good, three have to be better.
You and I may just have to agree to disagree on the above points, but I am sure you will agree with me that the forum readers need be familiar with the reasoning behind either approach.
rmw
