6 replies to this topic

[Stuka]

Hi,

 

I've been looking at single stage reciprocating air compressor calculations and I'm stumped by a problem.

Most sources I've found say that when dealing with the polytropic index for flow calculations for compressors, an ideal process is isothermal for the least amount of required work (n = 1). I know that n = 1.20 - 1.35 is the most common for this scenario.

Here is my problem:

1) In calculation there is a significant lower flow rate with n = 1 than an isentropic process (n = 1.41) but obviously less power requirement. Although at higher pressures, there is almost no flow with n = 1

2) In testing, having a compressor run for a while until it's hot on the exterior housing leads to a calculated n of ~0.8-0.9 (Indicating addition of heat into the compression process). Whereas cooling the housing down with air leads to n = 1.2 for higher pressures

 

 

So my questions are:

1) what is the ideal process? You get the most flow with n = 1.41, with some extra work required. Why would you want n = 1 to get almost no flow at higher pressures.

2) How does cooling/cooling fins really effect the polytropic index? I've found that when cooling down the compressor housing with blown air I calculate n = 1.2 (higher pressures), whereas letting the compressor get hot and without cooling it is around 0.9

 

 

[Bobby Strain]

Your query is puzzling. Reciprocating compressor is modeled as an adiabatic process, not polytropic. It is not unusual for compressors to have jackets around the cylinder to distribute the temperature, not for cooling.
What is affected by any cooling is simply removing heat during compression. And it is not characterized by changing n. Polytropic compression is a characteristic of centrifugal compressors.

 

Bobby

[Art Montemayor]

Bobby Strain is exactly correct.

 

You are “stumped” because you are relating to apples while dealing with oranges.  You can’t apply the rules of one to the other.

 

As Bobby succinctly states, reciprocating compressors operate under an adiabatic, isentropic process; centrifugal compressors operate as a polytropic machine.  You are trying to apply the principles of the latter to the former.  That won’t work for obvious thermodynamic reasons.  My responses to your two questions:

1. The ideal process for a reciprocating compressor is an isothermal one.  That is why multi-stage machines are more efficient than single stage ones.  Forget about applying any “n” (polytropic exponent) to a recip.
2. Cooling/cooling fins make no noticeable effect to the operation of a reciprocating compressor.  I know this to be an empirical fact because I’ve personally operated numerous machines with dry cooling jackets (I simply cut off the “cooling” water to the jackets) and never registered any difference in efficiency or operation during months of operation.  I did this on air, CO₂, and ammonia machines.  The answer is very simple: the cooling jacket is not a designed product.  All it is is a “best effort” attempt to cool the cylinders with what is available for a cooling surface ( which is small or next to nothing).  The best place to employ cooling on the machine (outside of the intercoolers) is the packing boxes or chambers.

I can’t imagine how you claim to calculate a credible “n” in a reciprocating compressor.  I’d like to see your calculations.

[Stuka]

How would I calculate the theoretical flow of a reciprocating compressor then?

[Bobby Strain]

You can go to checalc.com and find calculations. And use Google to search for information about such compressors.

 

Bobby

[Art Montemayor]

Stuka:

 

Calculating the theoretical flow of a reciprocating compressor is very easy.  All you need to know is:

• The diameter of the 1st Stage cylinder;
• Whether the 1st stage is single or double-acting;
• The stroke of the 1st stage piston;
• The rpm of the compressor;
• The volumetric efficiency of the 1st stage.
• The diameter of the 1st stage piston rod.

With this information - which you should readily have in your engineering files or in the actual compressor - you can easily calculate the expected delivery of the compressor.  I have designed the compression needs and resources for a lot of gas plants based on nothing more than that - and with nothing more than a slide rule.

 

I hope this helps you resolve your needs.

[breizh]

http://books.google....epage&q&f=false

On top of what experts suggested you , I will take sometime to read this book !

good luck.

Breizh

Edited by breizh, 02 October 2014 - 10:53 PM.