HOW PLATE HEAT EXCHANGERS BEAR THE THERMAL EXPANSIONS 2

[821326]

Hi all,
I am looking for some help on the ways of fabricating plate heat exchangers such that they remain undeformed at higher working temperatures than normal out door conditions; for example a common material used for HE is SS316H grade. this has a thremal expansion coefficient app. 17*10^-6 m/c. so for a HE of length arround 2m will experience a elongation of 30mm in a temperature difference arround 900 oC.

so there should be some fabrication tips to face this issue.
also if we are using a steel structure for the HE and use satinless steel plates to fabricate the tubes; then there can be a porblem, isn't it?


so i seek help.

thanks

8213265

 

[DYV1973]

When dealing with thermal expansion in a plate type heat exchanger, I would make sure that connections would be on 1 side. Due to temperature increase the HE will expand but this will not cause forces/stresses on e.g. piping. Just make sure that you do not have a rigid connection at both ends (1 side with rigid connection and pipe connections and 1 side should only be a support). But bear in mind that the expansion is of course in all directions.

What type of HE are you going to use (as you mention plates and tubes)?

DYV

 

[PHEs]

Some thoughts about it.
The mentioned temperature is extreme high for gasket plates.
For brazed plates,it is a block.
The temperature is a middle temperature of the temperatures.
All the plates do the same, follow etch other, except the endplates.
The endplates are normally plane plates.

With kind regards,

PHEs

 

[821326]

The HE will be a plate one where both hot side and cold side folws will pass through number of slots each of square section 1.5mm * 600mm at both hot and cold sides. lenght of these flow paths are app. 1.5m. that is why the expansion is critical.

 

[DYV1973]

@PHEs
900°C is a bit high for a gasketed PHE, and also seems a bit impossible for brazed types.

@821326
I think you are looking for some type of fully welded PHE, with corrugated plates or dimple plates. You can check on the Hybrid (from APV) or Compabloc (from Alfa Laval) for some inspiration.
What kind of application are you dealing with? If you do not need too much "thermal length", you can even look at the Plate&Shell types, as they have a corrugated plate pack in a fully welded shell.

PHEs is right that the plates will do more or less the same (if you are looking at parallel passages in 1 pass), but problem is that the complete PHE will expand.
So please explain the application.

 

[moltenmetal]

These conditions are beyond what brazed plates will be capable of. You need a printed circuit heat exchanger- basically a stack of plates diffusion welded together. Check out Heatric - Google will help you find them. Be aware, though, that even such an exchanger cannot handle an infinite amount of differential thermal stress!

 

[821326]

I am working on a design project of air to air heat transfer requirement where the heat is generated by diesel fired funace at a temperature arround 900 oC. and the load at the clod side to be maintained at app. 200 oC.
I selected the plate type heat exchanger because it is said to be highly efficient in heat transfer for air to air applications also i have been told to focuse on a plate type HE design.
I have already done the heat load calculations, overall heat transfer coefficients.... and the calculations with suitable safety factors shows a heat transfer are of app. 36 m2.
i have planned to use a one pass plate HE with welded structure and the plates to be inserted to the slots in the HE structure. the air flows are arround 200 m3/min in each flow.

my exact problem is

**what techniques to follow to fabricate the structure to avoid deformations caused by thermal expansions.
**if we use MS to fabricate the structure and use SS to fabricate the plate shells how to avoid deformations due to thermal expansion differences (is this a common practice in professional HE fabrications)
**or is wherther to use SS channels to fabricate the structure rather than MS..
** the lengh of the HE is to be 1.5m. this is due to the heat transfer area requirement(36 m2) where other two dimensions(in the transverse direction)are restricted due to existing duct sizes.
** also it is required to dismantle the plates or shells from the sturcture easily for maintenance needs in cases if leaks appear at the long time usage. so it is difficult use welded joints for fasten plates or shells to the structure.....

 

[PHEs]

OK 821326 you really mean 900 degrees C. At first i thought, you mean 90 degrees C.

At the starting up period, the varming up period, there is a problem with thermal expansion.
May be you can slowly heat up the MS, so that the materials can follow eats other, including the bolts.

May be I have a solution, but it is a long story with figures. It is possible for me to send it to your firm?

With kind regards,

PHEs

 

[rjw57]

I am not certain what you are trying to accomplish, but the temps you are talking about are not very practical in PHE's in my opinion. I have a lot of experience with broken brazes at slightly lower temps that you are talking about in several PHE designs (and even some fully wleded designs). Can you use a recuperator like this?...

http://www.wilsonturbopower.com/heat_exchanger/overview.php5

How about tube-in-tube like this?...

http://www.exergyllc.com/prod02.htm

 

[rmw]

While Wilson claims some very high efficiencies, what I see is a take off on Ljunstrom technology developed in 1906 which suffers from seal leakage rates that range from 8% guaranteed new right out of the box up to 25% after years of abusive service in hot gas streams.

So I suspect that while the wheel itself may be able to sustain the claimed efficiencies, the Hx as a whole can't or won't for long.

rmw

 

[rjw57]

rmw

I have no financial interest in Wilson's equipment so I am really only telling you what has been told to me. One of the reason's that this is perhaps a better design is that the unit indexes and recreates the seal after esch index. Therefore, there is no constant "scuffing" along with the corresponding wear. See this page, second paragraph...

http://www.wilsonturbopower.com/heat_exchanger/technology.php5

 

[rmw]

I considered that before I made my remarks. For the duration of the "lift" of the seal, the leakage is highest. Depending upon the rate of indexing of the wheel, the leakage should be pretty predictable. That assumes that the seal reseals perfectly and the leakage is limited to only the index duration. In a high temp application, I suspect the seals pay a price.

rmw