Freeware for calculating engine efficiency?

[Space car]

Could someone recommend freeware or a program of modest cost for evaluating engine performance? (Thermodynamics with heat transfer and friction?) I can't afford GE Power. I am trying to model a CI two-stroke.

 

[RodRico]

What specifically are your objectives with your CI two-stroke effort?

I am designing a compression ignition two-stroke (HCCI). I could find no model adequate to the task, so I used USC Professor Paul Ronney's spreadsheet "AirCycles4Recips.xls" to get started with my own model. I provided a link to the spreadsheet and his lectures (one of which focuses on AirCycles4Recips.xls) on an a post to this forum titled "Great Resource for fans of internal combustion engines..." on 5/6/2019.

The model itself is too course for actual engine design and lacks a good number of key features which I had to add...

1) The ratio of specific heats is fixed and relates to air only. You'll need to add a column for the ratio in each row of the quasi-static model and calculate it based on temperature and what's in the cylinders (air only or air and fuel or exhaust products).

2) Heat transfer is coarse. You'll need to incorporate convective heat transfer using one of the common heat transfer coefficients into the heat loss calculations. Google "______ heat transfer coefficient" filling the blanks with "Woschni" or "Assanis" or "Hohenberg" as appropriate for the design.

3) Ignition delay (critically important for compression ignition) is missing and needs to be added. There are many published papers on this topic.

4) Pumping losses are missing and need to be calculated from port flow analysis.

5) Friction is rolled up into a very coarse FMEP number. I use an averaged linear approximation of FMEP from measured data presented by Ricardo and Heywood (see below). The figures given on the plot are for 6-8 cylinder engines and analysis is usually done on a single cylinder, so I divide the average of the two by 6 to yield an approximate value.

6) There are too few rows of calculations in each portion of the cycle to approximate quasi-static conditions needed for valid results. Simply add sufficient rows to bring the number of degrees per step down to 0.1 or less.

7) Volume changes in a linear fashion during each portion of the cycle. You'll need to change it to match your crankshaft parameters.

8) The spreadsheet is for a four-stroke and you'll need to modify it for two-stroke.

Even the best thermodynamic models are only approximations of real world physics; models are only used to get in the ball park. If the model indicates success, the real engine may work but, if the model indicates failure, the real engine is likely a waste of time.

 

[Space car]

Thank you, RodRico,
I see you are ahead of me with your program. I have an air standard model with variable specific heat ratio, but no rational modeling of losses. We have both an analytical model and a finite element one. The finite element model can be extended to include losses, and your post is helpful. A challenge we have faced (in the FE model) is understanding the scavenging process. In general terms, our engine is a conventional two-stroke diesel with atkinson.

 

[RodRico]

Space Car,

When you said "CI two-stroke" in your original post, does the "CI" refer to "Compression Ignition" and, if so, are you referring to Diesel (well understood) or Homogeneous Charge Compression Ignition (far less understood)?

Do you have CFD capability? If so, it should be giving you some insight into scavenging. Solidworks CFD lets me work on mixtures of gasses, so I need only fill the cylinder with exhaust then push the scavenge air in and assess the mixture in the cylinder after the required time period. That being said, trying to get it optimized can be very difficult. Of all the scavenging methods, uni-flow scavenge in an opposed piston engine (like mine) is among the least complex, yet the videos at

http://achatespower.com/media-cente...of-a-two-stroke-opposed-piston-diesel-engine/

reveal characterizing it to be quite a challenge!

Rod

P.S. The software at

http://vannik.co.za/EngMod2TDetails.htm

is "only" $400 and may be worth a look.