Anybody here?

Could the easiest by - weld a lid / cap on it?

Grind end of tube to silver, put "lid" there - protruding "plug" you corner-joint fillet? - and weld MIG/GMAW at a lowish Amps (say 100A depending on tube thickness - so fuses to clean metal but very little penetration - so easy to sever when done?

Might be easiest... Not ideal solution if you are doing this 2 shifts 6 days a week 10 minute cycle time :-) but for a few-off, might be the readily done reliable one?

This page suggest a nickel seal might be up to the job 2200F is 1204C.

Grind open end and cover absolutely flat, then use annealed copper gasket between the two ?. Aka, cylinder headd gasket material.

Depends on how much vacuum leakage and absolute pressure is needed. You may be able to tolerate a small leakage if the right vac pump is used...

John

Of course, rethink this and use an inert gas rather then a vacuum ?...

Not a bad idea.

Would prefer not to have to use continuous pumping.

Still needs some sort of seal, as the regulator gives a few psi, but less stringent.

BTW if you are wondering why I don't want to use an over-insulation gas/vacuum containment, it's because a) it's expensive - a pressure cooker is the best solution I have found, and large tall pressure cookers are rare and expensive - and b) I want to use Kanthal A1 for the heating element, which needs an air/oxygen environment to maintain the Al_2O_3 coating.

Kanthal wire would be wrapped over an electrically insulating thermally conducing layer over the steel cross-section. Maybe glass tissue / ciment fondu or ceramic paper / filled 1700C HT cement for the insulating layer.

Type N thermocouple (like type K but better long-term stability) with PID controller - hmm, anyone got any ideas for a cheap programmable ramping PID?

Peter F

[1] they say DO NOT REFILL in big letters 0 but they have a handy refilling valve..

If you have 4/4" square cover, that's (correct me if i'm wrong), about

160 odd libs/sq" pressure across the face of the cover, which should seal pretty well.

Question is: do you really need high vacuum or not and if so, why ?. There will be some leakage, so continuous pumping will almost certainly be needed.

Need to define requirements a bit more tightly and fill in detail...

John

I was going to suggest an Arduino as there are a few PID temperature

module that supports a range of thermocouple types and has a support library but you can buy cheap Chinese PID controllers off ebay for

wouldn't bother making my own.

Copper as a gasket isn't going to cut it as it melts below your target

1150C temperature.

If you can use a circular opening then why not machine a taper seal? I wouldn't be surprised if one of the high temp anit-seize compounds would be good enough to lube the tapered surfaces. Eric

On 11/02/2021 23:21, David Billington wrote: [...]

claim to be.

IIRC so I

Peter

I wonder if it's like a friends PID controller which is a cheap Chinese? Brainchild? IIRC or similar, like

, where the controllers programming doesn't actually match what the documentation for the model says, looks somewhat similar to the cheap ones on ebay. Annoying in that the ramp rate is in degrees per minute rather than per hour which is more common in my experience.

Sounds like overkill to me, as if there's enough thermal mass in the system, the temp will ramp up and down slowly anyway, even using a standard on/off style controller. Don't make it more complex than it needs to be, even if writing the code on a micro is fun :-).

I usually look out for Omron 40mm EC5xx type controllers for such applications. Not too expensive s/hand on Ebay and some have direct dc control outputs to drive solid state relays, which avoids a mech relay...

John

No, they are just plain ordinary PID controllers with no process step functions at all.

price) for a cheap plain PID like the REX100, sells it as a programmable

Buyer probably notices it isn't a process control PID, but if not whoopee. If buyer notices, seller offers a 20% discount - if refused, in order to avoid the bother of shipping it back, seller will offer a 50% discount.

Maybe even offer a 70% discount - seller still makes a profit!!

Peter

I want process control - setpoints, soaks, ramps.

For instance, for investment castings I want to dewax for an hour or two at 90C, dry at a ramp from 100-250C over several hours, then ramp up to casting temperature of maybe 700C fairly fast, sound an alarm and hold 'til the melt is ready.

If I have to sit and watch over that I am not going to do it properly. A process control PID would do a much better job. And it would waste my precious time.

Don't make it more

I hate writing code. That's why I am a cryptologic theoretician not a cryptographic programmer.

Why? They are just plain ordinary controllers, no process control functions.

right model.

does a small part of the job.

Peter Fairbrother

Lol, if you had said to start with you needed a complex control function for this, then you would have got different answer. Virtusl project requirements drift, etc :-)...

John

The main - only, really - requirement is to be able to change the SV over time. A small extra requirement, for something which has the complexity of PID functionality already built-in.

Peter Fairbrother

Reciprocated.

Sorry - blind Bambi.