Finding a Bearing for 50Bar Pressure in Oven Heat

[Jack_O]

Hello, I am trying to find a bearing that will allow a shaft to pass into a loop with 50bar of pressure. The basic schematic is the idea since this thread:

The water jacket protects the bearing from high temperature as the oven will heat up the fluid to a maximum of 450dC. I assume I will be looking for a plain bearing rather than a rolling element bearing. Is there any of the shelf bearings that would fit the bill?

 

[Q_Goest]

Hi Jack_O. You have some fairly high pressure so you'll need to seal that around the shaft. So the bearing and seal can be in the hot area or they can be removed from that area, similar to what you show here, by providing a thermal gradient between the hot area and cooler one. The problem with using water in this case is that you need to seal the water and that water is up against the hot portion of the pump.

Can you simply extend the shaft to allow for a thermal gradient between the pump and the motor/bearing/seal? There might be some product inside that shaft area which may or may not pose a problem.

If you're looking for bearings that can take that much heat, since you may need something inside the hot area, I would think a bronze bushing would work if the RPM isn't too high and the service isn't too long. Any material is going to wear out fairly quickly in that service I would think. You'd probably prefer to use a plastic, but there's nothing I can think of that can handle 450 C. Lubricants are also a problem and can't be used, though you may find a dry lubricant that will work. Molydisulfide perhaps? Pure graphite?

Another consideration would be to either have a cantilevered beam or extend the shaft through the opposite end and provide a thermal gradient on the opposite side as well.

 

[Jack_O]

Thanks for your response, I was beginning to think this project is impossible. I've been in contact with a company called graphalloy which makes graphite bushings which can deal with high temperature. No water jacket would be required with a graphite bearing which could allow the shaft to be much shorter. The motor needs to be thermally isolated from the loop; it is likely that it will be connected to the shaft via a belt or chain, unlike the first diagram. In this case the motor will not support the shaft.

This is a more detailed idea for the shaft mount/bearing, my background is more physics than mechanical engineering though so I am not an expert in designing such things although it is an interesting thing to do. Any more advice welcome.

 

[Q_Goest]

The Graphalloy seems like a good choice. In general, your arrangement looks fine. I think the preference would be to put the bearings outside of the sealed area. At that point, you could even use rolling element bearings. If you drive the apparatus with a belt, you'll be putting a side load on the shaft which must be resisted by the bearings. Also, the seal you use will be affected by your bearing set up. The bushings that you show will alow some slop in the shaft, but seals don't like shafts that move around. Best to consider your seal and bearing as a combination.

Couple quick questions:
- What temperature does your material solidify at?
- What temperature do you need to keep your material above?
- Does the material require that you use any specific types of metal or is 304 SS acceptable?
- Do you already have the screw designed? If so, can you provide a drawing?
- What torque is required for the screw?
- What RPM is the screw running at?
- How long does it need to run for?
- Is this just for an experiment or is this a production unit?

 

[Jack_O]

I might be making a mistake here but I was imagining that the bushings would act as plain bearing and seal the loop at the same time. In the diagram I have shown exaggerated gaps between the bushings, I was think that the pressure would hold the circular bushings tight and sealed. As I say this is not my area of expertise.

- What temperature does your material solidify at?

This rig will mainly be for testing additives in a liquid call therminol VP1 which melts at 12dC and boils at 257dC under standard pressure (high pressure is to keep it liquid). VP1 decomposes above 400dC.- What temperature do you need to keep your material above?

Testing will generally be 350-400dC occasionally up to 450 with VP1. At a later date it may be used to test molten salts up to 600dC under ambient pressure, although a new similar rig may be built for this.

- Does the material require that you use any specific types of metal or is 304 SS acceptable?

Stainless steel is fine.

- Do you already have the screw designed? If so, can you provide a drawing?

Still unsure on the diameter and thread angle as the flow rate is not decided yet and graphalloy should provide the rpm limitation soon. Instead of fabricating a screw a suitably sized drill bit might be used instead.

- What torque is required for the screw?

Unsure as it depends on the resistance of the bushings.

- What RPM is the screw running at?

Under 600rpm, maybe even as low as 60rpm as the flow rate will not need to be large.

- How long does it need to run for?

Probably 10-20 hours per week.

- Is this just for an experiment or is this a production unit?

This is purely for testing, the design may be copied by a company for more testing in the future.

 

[Q_Goest]

Hi Jack_O. Thanks for responding to the questions. It helps to put your design into perspective. I see you're not worried about solidifying anything, the pressure and temperature are to maintain your fluid as a liquid, so we don't have to worry about heating the bearing or seal area. Those areas can be relatively cool so the fluid will be subcooled liquid in that area. If that's the case, I think you want to run your bearings and seal at ambient temperature by providing a thermal stand off between the hot process liquid and seals/bearing area. All you need to do to provide this thermal isolation is to extend your shaft away from the hot portion enough to allow for a thermal gradient between the two. Heat is rejected all along this thermal stand off either by using fins to reject heat directly to atmosphere, using water as you show in one diagram, or some other method. The clearance between the shaft and housing in this case needs to be as small as possible to prevent the liquid from setting up convective currents between the cold end and hot end. In the picture below, you can see a typical high temperature pump that has the motor, seals and bearings at one end and the hot impeller a distance away on the long thermal stand off. I would think your Graphalloy bushings would be ok in the hot end but regular bearings and seals would be added to the cold end of the pump housing.

Regarding the use of bushings as seals, that might work but the problem will be in maintaining the two rotating faces perfectly parallel to each other. If one side lifts up or even if it just has a reduction in contact pressure, the seal could leak. The way it's normally done is to provide a pressure energized seal mounted on a bellows or similar device that allows the seal to follow the face it is sealing against such as in the picture below.
http://www.conlyn.com/brands/pss/pss_components_diagram.gif
This image is for a seal on a boat shaft so it's very low pressure. The bellows especially isn't capable of handling any significant pressure. Better seals use metal bellows in this location, generally a type called "edge welded" to allow for the axial motion and to keep the seal against the hardened metal face. Spring load and pressure behind the seal help force the two parts together and create a seal. In your case, the flat washers you have might be able to accomplish this to some degree but they will need some preload, preferably by a spring pushing on the axis of the shaft, and they have to be absolutely flat and the shaft can't wobble at all or you'll have one edge lift up and it will leak.

Another option to your screw type pump would be a gear pump.
http://netpumps.com/images/gear_pump.gif
Again, you would still have a shaft that rotates inside a thermal stand off with seals and bushings at the end. Other options for off the shelf pumps might be found by searching ThomasNet.

Rather than making your own pump (or purchasing one) to circulate the HTF as a hot fluid, you might consider adding a heat exchanger to cool it off, pump it as a liquid at ambient temperature, then heat it back up to put back into your process. The heat exchanger can actually have the hot fluid coming into be cooled down exchanging heat with the cold fluid coming from the pump that has to be warmed up. This kind of counter flow heat exchanger is commonly used instead of operating the machine at high temperature.