Advice needed for metal repair

[Planapo]

Boys,

I just want to let you know about the outcome of our little "international workshop" on metal repair.

I've used a brand quality, slowly hardening epoxy resin adhesive (2 components have to be mixed together to obtain the usable glue).

I let the glue harden for 12 hours at room temperature, then heated the repaired breakage with a hairdrier for another 45 min., then let it cool and rest over night. So far it seems to hold. Stage and micrometer run very smoothly like the fine focus on a good microscope.



Thanks again, guys, for your help.

--Betty


PS and BTW: The insurance of the parcel service has refunded the price I had paid for the stage and the shipping costs.

[g4lab]

If I remember correctly the tensile strength of a good metal to metal
epoxy joint is on the order of 3200 pounds per square inch. (Pardon the non metric units) So if the shock loads experienced when something bumps it, are less than this the pieces should keep holding hands.

Nice looking job. The post cure helps.

Gene

[Harold Gough]

Come on, Rik, give us the calculations for the wattage of the hairdryer, distance, time, etc,. in relation to the heat capacity of the metal.

Harold

[rjlittlefield]

Sorry, I was distracted wondering why to heat the stuff at all.

Generally I try to avoid heat-stressing any glued joint, solely on the theory that metal and plastic are pretty much guaranteed to have different coefficients of expansion. But I have no reason to think that reasoning is actually correct or that it matters even if it is.

Can somebody more experienced with epoxies talk to me about the tradeoffs here?

--Rik

[g4lab]

The post cure increases the polymerization and bond strength particularly when the expected operating temperature will remain below the post curing temperature.

I think the joint stress is a non issue because during the post cure the viscosity of the epoxy drops and prevents forces from being developed that exceed its elastic limit. There is also probably an annealing effect. Cured epoxies are not crystalline but they are also not completely amorphous. The post cure probably improves the orientation of the molecules and probably reduces internal stresses that can develop in the cold cure.

The elasticity of epoxies is why I would have selected them over methcyanoacrylate glues which after curing are fairly brittle and would probably let go more easily if a shock load presented itself.

The first real job I had after graduating University was as a technician in a Materials Research Lab. The profs I worked for were up to their elbows in epoxy and epoxy composites.

[rjlittlefield]

Excellent -- thanks for the explanation!

--Rik

[Planapo]

You might like to hear about the little sequel to the story:

I had contacted Newport about a spare mounting block. They do have an office in Germany and I got a quote from there and the inch sizes of the hex wrenches I needed. Moreover, they informed me that I should have gotten a special key to assemble/disassemble the micrometer actuator with the stage when it had been originally delivered.

However, as the glue job was a success, in my reply mail I thanked them for their help and info, and told them that meanwhile I had helped myself with epoxy, but that now I would know where and what to order in case if something broke again.

The final email I got from the Newport people, at least one of them a physicist, would translate as follows:

Dear Mrs. ******

Thanks much for your feedback.
Even in todays era of high-tech the experimental physicist's most important tools are still the double-sided adhesive tape and the 2-component glue!

Have a nice weekend.

Kind regards,

Now, wasn’t that nice?!

--Betty

[Harold Gough]

The elasticity of epoxies is why I would have selected them over methcyanoacrylate glues which after curing are fairly brittle and would probably let go more easily if a shock load presented itself.

My (domestic) experience is that they do let go.

Harold

[g4lab]

I have had much more experience with superglue letting go. This is not to say that I don't like and use superglue. But I believe Betty's choice of epoxy is correct.

Any joint is going to let go if its bond strength is exceeded by the loads presented. Shock loads are typically in multiples of the usual loads and that is almost always what causes failure.

Often these failures occur because the bond was not good in the first place.
On metal this means cleaning the surfaces with a really clean solvent that has no oil residue. Sometime paint store solvents have enough oil in them to cause bond failure. Before you use them as a cleaner you have to do the glass plate test. Let some evaporate on a clean glass plate or watch glass and see the amount and character of any residue that is left.

If any silicone oil has ever gotten near the surfaces you can forget about bonding them. Some spray lubricants have silicones. WD40 and Liquid Wrench do not.

If you use (as seems to be the case domestically very often) "Five Minute Epoxy" or other fast cure epoxies you should know that the strength they develop is about half of "real epoxies". So avoid those.

If you have a balance sensitive enough to weigh the small quantities usually needed for domestic repairs , you should utilize it for accurate weighing. Usually the formula calls for a 50/50 mix of hardener and resin.
There are variations in formulae though and the manufacturers directions should be consulted and followed.

After accurate weighing out of the components we would subject them to a vacuum for a while to draw out as much air as possible. Sometimes while heating the components (separately) to lower their viscosity. We would then combine them in a vacuum flask with a glass stirrer with teflon blades and stir in vacuo for a period of time. Then sometimes there was a period of reaction allowed before specimens were made. There was almost always a high temp post cure. Sometimes we cooled the flask to remove exothermic heat and slow the cure rate down. The thermal history makes a difference. If the resin component has been on the shelf for a long time and gotten semisolid it can be "remelted" before use which will aid in mixing and reacting.

Sometimes you can solvate the components a bit to promote complete mixing. A very fast evaporating solvent should be used. Acetone or MIBK can be used if it is fresh and dry and you live in the desert where it won't pick up water. Hexane would probably be better. There is a pretty clean artists version called Bestine that seems to evaporate pretty cleanly.

By adding more hardener you will get a more brittle and faster curing epoxy. If you add excess resin the result will be a bit more flexible and elastic.

There has to be enough surface area in the joint for everything to keep holding hands. The few times I have had epoxy fail it was because the surface area wasn't enough or the effective surface area wasn't enough because of contamination.

[Charles Krebs]

g4lab,

After accurate weighing out of the components we would subject them to a vacuum for a while to draw out as much air as possible. Sometimes while heating the components (separately) to lower their viscosity. We would then combine them in a vacuum flask with a glass stirrer with teflon blades and stir in vacuo for a period of time.

Wow! ... I use a toothpick and a piece of scrap paper...

Actually, on occasion, I'll use that "negative" characteristic of superglue when I want two flat surfaces held together strongly, but know I will be separating the pieces after some use. A single-edge razor blade is placed against an edge of the joint, and the back tapped sharply with something hard.

Are there any epoxy "brands" you find superior? (Especially ones that are readily available in stores. I'm sort of partial to the JB Weld mentioned earlier in this thread).

[Planapo]

I've used "Uhu endfest 300"
http://us.vario-helicopter.biz/shop/pro ... s_id=33436
http://www.uhu.com/_uk/produkte/2_kompo ... dfest.html

which is also recommended and used by professionals and for large scale commercial applications.

I now read in the data sheet that I've just found ( http://www.supermagnete.de/docs/uhu_plu ... 00_eng.pdf ) that the maximum strength is achieved when it is hardened for 5 min at 150-180 °C. However, in this case I would not have wanted to put the micrometer that was fixed in place all the time, under too much thermal stress, and even a bond strength of 1200 N/cm² should be more than enough anyway.

--Betty

[rjlittlefield]

That's a very informative data sheet -- thanks for the link!

--Rik

[g4lab]

The UHU product looks very good. Since they have taken the trouble to test and post a detailed application sheet I would not hesitate to use it at all.

I also use JB Weld.
I have used 3M Professional products epoxies. Also Loctite products.

I use a patch of aluminum foil and a wood Q tip handle.

The products we used at the MRL were made by Shell Oil Chemicals division and others. They usually came in a drum. When we were done mixing them up we would remove the stirrer and take its teflon leaves off and throw those into some strong solvent. The flask and glass stirrer rod went into a heat treat oven and were taken up to dull red (in a hood of course) until about an hour after the smoking stopped. They would come out spotless. That always tickled me.

[Harold Gough]

A single-edge razor blade is placed against an edge of the joint, and the back tapped sharply with something hard.

Do you practice the technique by cutting diamonds?

Harold