Page 3 of 4

PostPosted: Sat May 04, 2013 7:45 am
by GreenBean
I hope you enjoyed your holiday before your luck ran out again. :( Did you clean all the coffee oils off the rear wall before applying the JB Weld? From memory I think I am correct in saying the JB Weld instructions advise that you should use pure acetone to clean the surfaces and must not use any other petroleum based product as this will make the bond fail. I suggest you clean it to remove any coffee oils, then sand it to roughen the surface, then clean it a few times with acetone being careful not to wipe any coffee oils from the surrounding area.

[Edit: added a copy of the relevant FAQ from the JB Weld website.]

PostPosted: Sat May 04, 2013 9:31 am
by silron1
Yep - I used acetone with great care but it did clean the wall surface prior to fixing my TC holders on the rear wall.

I used the Dremel to roughen the metal surface prior to fixing with the JD Weld. I had a machine apart a week or so ago & the weld appears to be holding up very well.

Ron

PostPosted: Sat May 04, 2013 3:06 pm
by frcn
I used some emery cloth to get though the surface "treatment" whatever it is and to roughen the surface and then used lacquer thinner until the rags came up clean. Have not had a problem with the few I have done. it was actually difficult to clean off any "spread" ont eh epoxy that impinged into unwanted areas.

PostPosted: Sun May 05, 2013 11:21 pm
by simonp
Just read the advice here after re-doing it :roll: Luckily though I had ued a dremmel to roughen up the surface of the wall and the spacers, and also cleaned thoroughly.

The bad news is I have just done another roast and one has come adrift again :cry: It is too hot to investigate inside yet. I heard the joint crack at about 180C envireonment temp.
To add to this the readings were again way too low with 1st crack coming at an indicated 180C from the bean mass thermocouple. So far this has been a complete disaster and I don't think I am going to be able to use the TC4C & HTC boards at all at this rate.

I defifntely followed all the correct procedures for the JB Weld, so either there is something odd about the stainless of my backwall, I am getting contamination whilst fitting, or there is some usdue side stress on the spacers when the thing is running.

PostPosted: Mon May 06, 2013 12:03 am
by simonp
Well, both had gone again, the environment temp one completely loose and the bean one with cracked adhesive but still attached. See photos below:

Image

Image

PostPosted: Mon May 06, 2013 7:22 am
by silron1
What a shame, you really seem unlucky with your upgrade.

As I said earlier I have used JD Weld on my mods without any heat problems.

I did fix the BT & ET probes through drilled out rivet nuts welded to the rear wall behind the roasting chamber. I then wrapped some Kapton Tape around the collars so that it kept the probes adjustable & appeared to ease any strain on the fixings. I posted full info on this much earlier in this thread.

I hope it works out for you.

Ron

BTW - I seem to be using much more Weld than you. I end up with it very thick around my rivet nuts & then cure it overnight under a strong table lamp.

PostPosted: Mon May 06, 2013 8:53 am
by GreenBean
As you say, it does appear from your photos that it is the bond to the rear wall that is failing.:( I still think the most likely reason for that is contamination of the surface. What substance are you using for the initial and final clean of the rear wall surface? If acetone is it pure acetone? Is it completely dry when you apply the JB Weld? Is it fully cured before you roast with it?

The thermocouple reading low at first crack start is not a significant problem. The default RoastLogger settings only need one change to adjust for that. The default settings assume a reading of about 190 C at first crack and make the first turn down of the heat at 178 C. If your readings at first crack start are 180 C (10 C lower) then simply lower the first turn down temperature in your initial roasts by 10 C to compensate as shown in the image below. You can then fine tune your settings as described in the RoastLogger manual.

PostPosted: Mon May 06, 2013 10:51 am
by bruceb
I've done a good bit of metal bonding using various products and from the picture above I would also agree that the surface was 1) not completely free of grease and/or other contaminants and 2) too smooth. I don't know if it is possible or makes sense, but a couple of small holes to the left and right or top and bottom or both of the thermocouple so that the J-B Weld can go through and harden on the other side might be worth considering. If not holes at least rough indents.

PostPosted: Mon May 06, 2013 11:36 pm
by simonp
So, third time lucky?

I had previously cleaned with soap and water and the surface of the metal had been roughened with a metal end on the Dremmel. This time I used a grinding stone and also scratched with a needle file and used IPA cleaner (which claims to leave no residue at all). We'll see what happens.

Curiously I used the same roughening on the stainless spacers as on the back-wall and the adhesive stuck to those fine enough. I wonder if I am getting some undue flexing of the back-wall under heat which the JB Weld (being quite rigid) cannot cope with?

PostPosted: Tue May 07, 2013 8:06 am
by GreenBean
The main stress on the joint, and hence the bond to the wall, is due to the differential thermal expansion of the hot wall and the relatively cold spacer supporting the thermocouple. A proper bond can cope with this stress and also stiffens the rear wall in the vicinity of the spacer.

Other people have succeeded in using JB Weld for this purpose without the problems you are having. If we consider the differences between your approach and theirs you are using the same product (JB Weld) and have roughened the surface in the same way. The main difference between approaches is that others have followed the JB Weld instructions and used acetone or cellulose lacquer thinner and you have used IPA.

The JB Weld instructions (in a post above) specifically state "DO NOT use alcohol or any cleaner that will leave a petroleum residue" . In order to break the cycle of failed attempts I would suggest that you abandon the IPA and follow the instructions to use acetone or cellulose thinners. :wink:

PostPosted: Tue May 07, 2013 8:26 am
by simonp
GreenBean wrote:The JB Weld instructions (in a post above) specifically state "DO NOT use alcohol or any cleaner that will leave a petroleum residue" . In order to break the cycle of failed attempts I would suggest that you abandon the IPA and follow the instructions to use acetone or cellulose thinners. :wink:


Hmm, I suppose there are 2 ways of reading that instruction. I had read it as "do not use alcohol/cleaner that will leave a residue" and I am using alcohol that claims not to leave one. I guess you are saying it means "do not use alcohol (of any type), or a cleaner that will leave a residue? A subtle but perhaps important difference!

This is actually the FIRST time I have used the IPA, previously I used soap (well washing up liquid) and water.

Hmm, looking around it looks like IPA does leave some residue (despite what it says on the tin) so I suspect I will have more issues:roll:

I don't have any acetone but I guess I need to source some, I suspect the wife's nail varnish remover is not very pure :wink:

As you rightly say that seems to be the only real difference in what I have done, so logically it must be the problem.

PostPosted: Tue May 07, 2013 8:30 am
by simonp
Ah, hand on, from the www.jbweld.eu website
" Completely degrease the surface (eg with isopropanol, or acetone) "

IPA being isopropanol should be OK. We'll have to wait and see I guess.

PostPosted: Tue May 07, 2013 2:12 pm
by bruceb
In my opinion no alcohol or detergent will remove grease as well as acetone. Isopropyl alcohol is, in itself, a good solvent, but it is hygroscopic and dilutes itself so that its ability to dissolve non-polar (lipid) molecules becomes limited. It can be dessicated by adding salt, causing the water to become "salted out," but this is far more trouble than picking up a small can of acetone at the local building supply or hardware store and using that.

PostPosted: Tue May 07, 2013 3:19 pm
by simonp
bruceb wrote:In my opinion no alcohol or detergent will remove grease as well as acetone. Isopropyl alcohol is, in itself, a good solvent, but it is hygroscopic and dilutes itself so that its ability to dissolve non-polar (lipid) molecules becomes limited. It can be dessicated by adding salt, causing the water to become "salted out," but this is far more trouble than picking up a small can of acetone at the local building supply or hardware store and using that.


Fair enough, I guess my solvent knowledge is quite limited, I should stick to electronics :wink:

I found a nice bottle of dichloromethane in the cupboard at work which is a polar aprotic solvent - according to Wikipedia :lol: - like acetone and will leave no residue, so if I have another failure I will try that.

It is a shame the european and USA websites have different advice with regards to cleaning preparation :roll:

PostPosted: Wed May 08, 2013 8:34 am
by simonp
I did a bit of an experiment last night, I tried just running through a pre-heat to see what happened with the outer cover on but not that back panel. I ran it till about 120C on the on-board temp sensor. This reached 150C during the cooling cycle. I haven't taken off the shield yet, but peering in the back there don;t appear to be any cracks in the JB Weld but it has (as before) changed colour after the heat from a a dark grey to a light-ish one.
I might try a full run with no beans and see what happens. I am wondering if the beans are putting a side load on the thermocouples which is in turn levering the spacers causing the problem? Maybe a slightly flexible end on the thermocouple gives an easier time?