Model Engine Maker
Help! => Hints, Tips & Tricks => Topic started by: Zephyrin on July 10, 2018, 11:35:01 AM

Hi,
I recently completed an Excel spreadsheet for calculating the power of vacuum engines or flame gulpers.
This model is based on the evolution of the temperature of the hot air in the cylinder during the cycle, and the resulting pressure, according to perfect gas laws.
Calculations follow this line:
A mass of hot gas enters into the cylinder during each displacement of the piston (corresponding to a crank rotation of 5° in the table), mixes with the gas already in the cylinder, the temp of the mix is calculated, as is the heat loss by heat transfer, and the subtraction of corresponding Kelvin (Cp=1000) gives the new internal T for the next step.
For the exchange of heat, the surface of the cylinder at each step plus that of the piston are taken into account, not the cylinder head, constantly heated by the flame…
The pressures variations that may result of the progressive opening and closing are set aside, as they depend on the valve kinetics of each engine.
The adiabatic processes occur after the closure of admission, and are taken into account in addition to the above thermal exchanges to calculate the pressure in the cylinder during the cycle.
The work done is calculated at the end of the cycle, as is the power.
As expected from vacuum engines, the power figures are low, and one had to subtract the mechanical drags and pressure losses, (the admission doesn't close instantaneously etc...). The power returned by this model relies largely on hypothetical parameters as flame temperature and heat transfer coefficient, but is not far from the values I have measured on a real engine (already shown on this forum) with a brake dynamometer and a tachometer! Instant gas temperatures measurements are not accessible in the home workshop, which is not the case for the pressure monitoring in the cylinder during the cycle, a PV diagram should be experimentally feasible, to validate this model, I guess....

Hi Zephryn, that is an impressive spreadsheet. I must admit that I tend to stop at the mechanical dynamics and stop short of trying to calculate power, but it looks like a really good attempt. It even downloads nicely into Numbers on the iPad.
As you say, a lot depends on the actual temperatures and heat transfer coefficients, but the delightful thing about spreadsheets is the ability to do some trial and error, so we at least learn how important the items are.
With such a spreadsheet, I generally make the first column with a formula so each line is increased by a set amount which I put in a cell near the top. Then, once the formula are all complete, and it is all working, I can reduce the value to say 1 degree intervals, and then copy the formula down the necessary extra lines, which sometimes makes things go a bit smoother of the coarser interval makes things jump a bit. Other times it is not necessary.
I am also interested in your dynamometer, they seem to be the flavour of the week. Could you please post a little detail on your dynamometer.
A great spreadsheet, thank you.
MJM460

Thanks a lot for sharing your spreadsheet :atcomputer:
That's very interesting!
Plani

Thanks for appreciations !
MJM460 : I'm not familiar with spreadsheets, I got the point how to facilitate the use of formula, thanks...
Keeping the number of column manageable leads to formula so complex that I'm rapidly out of touch! Often I regret Qbasic...
I will post tomorrow pictures of the device I use to estimate the power of my engines, as you will see, tinkering with bits and shoestrings, far from the hightech setup that Plani has made, shown in his thread !

Hi Zephyrin, for those calculations, spreadsheets are your friend. For someone not familiar, you have done pretty well. It is really interesting and informative to explore the engine performance in that way.
The biggest issue is to set them out so they can be checked. Complex formula which extend over several lines in the formula bar are quite difficult to check, but you are nowhere near the limit yet.
Don't be shy about using plenty of columns. When I was doing big spreadsheets at work, the right hand side limit was column IV, or 256 columns, which I used up many times over. Most people who discover the limit are using the spreadsheet as a data base, but mine were all formula in the style you are doing. I had to repeat the results from the first sheet into the first few columns of a new sheet, then continue on. However the later versions of Excell accept 16,000 columns, I suspect it might be 16 k or a little over 16000. Will normally be enough to avoid overflowing. But I know from discussions with my colleague who had to check my spreadsheets, that it was easier for him if I used extra columns to calculate each of the factors, then used a simple formula such as your "if" or lookup or logic formulae without further imbedded calculations. You won't run out of columns these days.
The other challenge is to document the calculations clearly so you can still follow it a year or so later. You seem to have done well on that front. If you have trouble with some formula, send me a pm, I can probably help. Adding the formula is relatively easy with practice, it is knowing which formula to use that is the hard part.
It is good to see someone else exploring some of the real power of spreadsheets. It is not practical to do those calculations without a spreadsheet.
MJM460

Hi,
thanks for all these advices...
I have changed the increment in the table to 1° instead of 5°. Changes in the power are barely visible, as area are more precise.
I don't want to go deeper in the calculations, it was simply to tests a possible model for vacuum engines, similarly to the many spreadsheets that exist for Stirling engines.
I was expecting a different profile to this that emerged from my sheet; a larger cooling through exchange at BDC and drop in pressure, isochoric, owing to the time spent by the piston at BCD, but it is not so...
But I agree, it is a useful exercise, not only for my gray cells, I see clearly that the valve settings have to be precisely controlled even if the exact point of closure or opening are not that easy to determine !