Heater Design Advice 2

[Rmechanical]

I need to design and come up with the analysis of an oven, but just a few guidelines were given to me. I'm a recent grad so I'd like some advice on this because I've never done something like this before. Here are a few guidelines:

- The oven is cylindrical. About 5m long and 20 in. diameter

- The heater is at one end, and the temperature must be maintained at 170C within +- 1 degree across the whole length. (a fan would be used to circulate the hot air and the exhaust will be fed into the system - closed loop, and the flow shall be laminar)

- Inside the main chamber, another sub-chamber will be built. The sub-chamber will have an inner and outer material of different thermal conductivities to allow a uniform heat transfer to the area where the sample will be tested.

- The oven will be insulated, so the walls can be considered adiabatic to the outside. Pressure is held constant as well.

I need to calculate the temperature distribution in the inner chamber every 2 feet, come up with the dimensions for the heater (length, wall thickness, etc), the power needed for the system, materials, the time needed to heat the samples, and make sure the temperature in the testing area is maintained constant at 170 C.

I'm unsure in how to tackle the problem, and we don't have a senior engineer to give us advice, so I'm on my own.

So far I've used the formulas for convection and conduction heat transfer, but I'm having problems calculating the heat transfer coefficient, h, for air. I've been using the Reynolds and Nusselt numbers, but the values I get for h are <1, and I read the values for air are typically from 1 to 100, and even more in forced convection.

I've also looked into formulas for extended surfaces from my heat transfer book in order to get an approximate heat distribution.

In order to see how much time is required for the sample to heat up, I've been reading the lumped capacitance method, and to calculate how much energy is needed to heat everything from 25 to 170C, I used the heat capacity and mass of each material to solve for energy in joules.

My numbers seem quite realistic, but since I have no previous experience in the matter I'd like to revise them.

Any advice would be greatly appreciated.

Thanks in advance.

Roger

 

[Compositepro]

You are getting way too theoretical. Oven design is mainly empirical. Keep in mind that if there is no heat load there is no temperature difference and if there is heat transfer occurring there must be temperature differences. Air has very low heat capacity so in order to convey much heat energy its temperature must change significantly or there has to be very high velocity to get high mass flow.

In most ovens, much of the energy used is heat loss though the insulation so it is generally not a good idea to neglect it.

When you are heating up to temperature you are trying to pump heat into an object and you cannot have a uniform temperature. The faster you heat the less uniform the temperature. If you are trying to maintain a uniform temperature you must minimize heat loads and losses. You cannot have it both ways.

I would guess that an oven your size will need a blower of at least one horsepower. Blower horsepower adds heat to ovens. There are ovens that go over 400F that use only blower power as a heat source.

There are also many ovens that don't have blowers that can hold fairly even temperatures. These work fine for keeping thing warm but are very poor at heating things up quickly.

The key to uniform temperature is high air velocity and this means turbulent, not laminar flow. In a long oven as you describe air will flow along the surface of an object and this is bad for heat transfer. Air must impinge on a surface and immediately be flushed away to transfer heat. So crossflow would give better uniformity.

 

[Rmechanical]

Thanks, it was really helpful. Maybe I'm complicating things too much, but I wanted the design to be backed up by the math behind it, and to have proper documentation.

You were right. I calculated the heat loss going through the insulation and it was significant. I assumed a heat transfer coefficient for air, but I'd like to calculated or at least have some values relevant to this particular application. Can I get some help on this?

For the cross-flow, I think that won't be possible since we can't have any metallic parts parallel to the testing area (metal can be used at the ends only). And the thermal conductivity for non-metallic materials is quite low.

I'll make some calculations using a higher air speed as well. The person who told me about the design said he wanted laminar flow, but the speed of the fan/blower needed to be way to low; since it is not a critical part of the design I'll go with turbulent flow.

I'll be attaching a simple diagram later so I can explain it better.

 

[gwolf2]

Rmechanical,

Well, you've really been thrown in at the deep end!!

The htc in this situation is really difficult to get right. With such a long oven the boundary condition growth and laminar/turbulent transition distance along the shell are horrible to calculate. I agree with compositepro - try to force turbulent flow via high speed to at least dump it into a known flow regime.

I have had quite good results with this type of problem by using a spreadsheet program and dividing the area into sections. I lay out a sketch of the problem with line drawings in excel and then start putting properties and relationships into boxes in the relevant location. I then hook the boxes up with relationships, assign some guess temperatures and iterate. I usually use a very simple vba program to iterate temperatures based on net heat flow into/out of each box and increase or decrease the temperature by a small ammount accordingly. It´s a quasi-transient approach and really quite simple.

If you did go this way then treat any metal bits as uniform temperature heat sources/sinks because the metal conductivities are so much higher than the htc´s that you can get convergence problems.

gwolf

 

[Rmechanical]

Thanks gwolf. Yes, I feel like I was thrown into a race right after I learned how to walk! But I love the challenge though.

Yes, finding the htc has been the most difficult part. At this point, I will settle for an approximate value. I'd like to make an assumption on that, but since I have no previous experience doing this, that's a bit of a problem.

Right now I'm trying to learn FloWorks. We have it at the office but I'm new to Solidworks. I turned out to be good at modeling and stress analysis with COSMOS but FloWorks is a different story.

I'm attaching a simple sketch so you guys can visualize the oven a bit easier. Let me know your thoughts in the design.

Roger

 

[Compositepro]

Normally the blower would be in one end because the motor is not in the heated space of the oven. The heated air should flow to the other end and return over your testing chamber. This gives maximum opportunity for the heated air to mix and reach a uniform temperature. It not clear form your drawing but you should have one single insulated envelope and uninsulated sheet metal inside to direct air flow. With this arrangement you will definitely have one end of the test chamber heat-up before the other end. The only way to minimize the temperature difference is to heat-up slowly. Once at temperature there should be no problem if there is no on-going heat load.

You can put your thermocouple in the air after the heater, just before the test chamber, to prevent overheating.

 

[Rmechanical]

Yes, that's true. The blower needs to be outside, and I like your idea of having the hot air coming through the sides and returning over the testing chamber.

I also agree that sheet metal would be ideal to allow a good heat transfer, but I can't use any metal because the readings of the sensors in the testing specimen would be affected by it. No metal is one of the main requirements.

I've been reading but I can't find god non-metallic thermal conductors, so if anyone can provide any insight in this subject I'll be very thankful.

Thanks a lot to the people who has responded, especially Compositepro. You have helped me greatly!

 

[chicopee]

Don't forget to interlock the heating system with the fan. If the fan breaks down the heating system must shutdown, otherwise you run the risk of burning up the heating filaments and also create a hot spot in the oven.