High Temperature Alumina Reactor

Hello fsmonder,

I have a vague idea of your setup, but it is missing some crucial data. If I am able to help, it would help immensely if you could provide a sketch of the reactor, including the orientation of the cooling at the top end of the reactor. If you can't give that, fine, but here are some questions that need to be answered before a cogent reply can be given:

1) Is the water-cooled section integral to the reactor or is it a jacket?

2) Is the reactor itself insulated?

3) Why does the top of the reactor need to be cooled? Can the reaction mixture not be cooled once it has left the reactor?

With cooling at one end of the reactor, and at these temperatures, the obvious answer is that the cause of the tube's cracking is dissimilarity in expansion/contraction between the cooled and non-cooled zones.

Please reply,

Mike

Dear Sir,

Thank you for the reply. I will be very glad to email you all details of the system.

Kindly let me know your email address. Answers to your points-

1. Water cooled section is integral/cemented to the reactor top. The TOP is a

flanged joint. The top closing flange has water channel routed through. It has openings for four R type Thermocouples,that go into the reaction zone. There

is another opening for charging in the material and also taking a sample of the

reactants. These details will be sent. The lower flange of the top has a 4" tube

section, that is double walled for water circulation. (We will stop this part cooling)

2. Above the furnace top, there is 6" insulation. Inside the reactor, it is open upto

the top. Do you think we should put an insulation barrier inside that bars free flow

of heat -radiative and convective-to the top?

3. The reactor top flange can not be exposed to high temp. Thermocouple junctions

are nearby. The swagelok ferrule fittings and nuts cannot take higher temp.

Yes, you are right about the reason for the cracking.

I await your email address. Thank you once again.

fsmonder

Hi,

cracking is likely either by

lengthwise temperature gradient (aggravated by roughness and micro-cracks)

or by

lengthwise stress or bending stress from the flanges and clamping the flanges to the frame.

Ramping up and down should not be a problem.

Is the tube used as fired (better in finish?) or is inner or outer grinding done to get tighter tolerances?

I would try to lengthen and smooth the zone where the temperature gradient exists (As you intend to do).

Alumina has a pretty good thermal conductivity, so you will have to balance thermal tube insulation of the upper section giving radial heat loss with thermal conductivity of the tube itself in order to get the wanted axial temperature profile.

Quartz-glass may be easier? Lower TCE and much lower thermal conductivity.

RHABE

Sir,

Yes you are right, the cracking is due to axial temp gradient.

We had the entire top length of the tube in insulation, and there was heat leakage

from top zone. So, the temp gradient close to the top joint became steep. Water

cooling will be stopped here, and the tube outer insulation removed. Also we

will put an alumina foam plug inside the tube above top zone, so no direct

radiative heat transfer takes place.

But the question is, what length of the bare tube above top zone will make a safe

temp gradient possible.

I am also trying to get KOVAR, to make a seal between ceramic and SS304 flange.

The tube is not ground, and is used as fired; but it was very smooth and fine.

Will report back progress. Thanks very much for the inputs.

fsmonder

Your problem is the differential thermal contraction between the stainless steel and the alumina. The top joint should be flexible or you should select a transition material for the section between the top of the alumina and the stainless steel that matches the thermal contraction of the alumina. Usually Invar or Kovar is used for such joints. In any event, the alumina tube should be in compression at any joint because alumina is good in compression and poor in tension. Next the transition material can be welded to the stainless steel. The joint between the alumina and Kovar can or should be a nickel braze.

If the cooled flange is cool enough to use a gasket, the alumina can be connected to the stainless steel flange using a gasketed joint using Grafoil or any gasket that is compatible with the process fluid and suitable for the temperature.

I can be reached at ability@ameritech.net.

Sir,

Your comments are very insightful. I will try to get KOVAR sheet,bend it into a

tubular shape, and braze (Nickel) alumina tube to it. Might be able to salvage

the current broken tube- only 2.5" length is broken.

Gasketted joint is possible ,but the tube ovality is a limiting factor. I will also try

a jointing putty ( AREMCO, USA- pyroputty950) that we have .

Thanks for advice, will report back.

fsmonder

Subsequently, we have made another try.

The cracked end was machined off: this decreased tube length by 6" to 33.5".

At upper end, we cemented a 4" length of Kovar tube and at lower end, we

cemented at 2.5" long Kovar tube. At upper tube end, we put INTERNAL insulation (ceramic board4"thick ) through which the four internal thermocouples passed loosely, allowing gas flow.

Externally, above top hot zone we put 2" (1600C rated) board and another 2" 1200

rated board. Ramp up to 800C at 10 C/min, to 1000C 4C/min, to 1450 at 3C/min.

Cooling down, same ramping was used.

THIS TIME THE TUBE CRACKED ABOUT 2" ABOVE THE TOP HOT ZONE. It is

a circumferential & neat crack. The Kovar section starts about 4" above the TOPHOT zone. At the start of Kovar temp was about 600C. Over a 4" tube length, temp drops by 800C. Is it too high a gradient?

We are trying to get new tubes, but the scene is getting very scarry. Pl. help.

Can I email you a few pictures?

fsmonder

Are you sure your problem is a thermal contraction problem? We have been assuming it is, but is it? It may not be a stress problem at all. Maybe the material can't tolerate the gradient. What does the ceramic manufacturer say about the high gradient? Did you look at the differential contraction between the two materials?

Your idea of insulating the end and bringing the gasses in and out of the vessel through the insulated deadman is a good approach.

I have seen ceramic tube reactors placed in a containment vessel with an o-ring seal at the top and bottom. These were radial o-rings which sealed on the tube OD. That is a reasonably low stress connetion and could be quite effective. If you keep the length long enough to keep the seals cool, you could do a variation of that design without the containment. You rely on the ceramic tube as a pressure barrier as it is so it isn't too much of a stretch.

I can be reached or contacted at ability@ameritech.net if you need to discuss further.

Kovar thermal expansion date CTE (Carpenter) 4.90 µm/m-°C @Temperature 30.0 - 400 °C 5.13 µm/m-°C @Temperature 25.0 - 300 °C 5.30 µm/m-°C @Temperature 30.0 - 450 °C 5.86 µm/m-°C @Temperature 25.0 - 100 °C 6.15 µm/m-°C @Temperature 25.0 - 500 °C 11.26 µm/m-°C@Temperature 25.0 - 900 °C Alumina: (Coors) 8.10 µm/m-°C@Temperature 25.0 - 1000 °C 8.10 µm/m-°C@Temperature 25.0 - 1000 °C 8.10 µm/m-°C@Temperature 25.0 - 1000 °C 8.10 µm/m-°C@Temperature 25.0 - 1000 °C Hi, as to be seen from above data Kovar is a good choice for temperatutes up to 500°C. If you need an endcap made from metal, then a thin section is advisable so that any difference in diameter and slope with temperature can elastically or plastically deform this thin section. I think that you cracked your tube with the lengthwise thermal gradient. I can calculate the allowable thermal gradient if you supply geometry and material data, but this is not an action done in a few minutes. If you buy new tubes be very careful with packing and handling, any tiny scratch (not seen by naked eye) - especially if circumferential - will lower the strength considerably. Be also sure to get "as fired" tubes. Grinding is not at all recommended for these highly stressed applications. Pasting fotos is a good idea. RHABE