I put together an excel spreadsheet that calculates the proper turbo compressor air-flow for any given engine and pressure ratio. It also calculates displacement based on bore, stroke and # of cylinders, and finds hp, tq or rpm given any two of the values. You can open/download it here. Any cell that is filled with grey is an input cell and should be updated with the specifics of the engine in question. The thing seems to be right on for the most part, but I have a few concerns.

First, I used a rule of thumb found in hot rod's Turbo's, Part 1 article that every #/min of airflow is worth about ten horsepower. I used this for estimating horsepower at peak rpm and at two other relevant points. I realize that this is not part of actually choosing a turbo, but it helps the spreadsheet estimate real world results. Is this 10 hp for every #/min assumption valid, and/or correctly applied here?

Second, I then applied the hp = tq x (rpm/5250) equation to yield torque numbers for the same three points. The tq numbers turned out to be the exact same at each point. I am quite sure that the equation is entered correctly, so I feel that my rpm vs. airflow assumptions are incorrect. If you notice, I used this for full flow rate at full rpm, 50% flow rate at 50% rpm, and 20% in the same manner. First, are these good points to plot on a compressor map to determine the suitability of a given turbo for a given engine? And secondly, does rpm and airflow vary linearly as assumed here? It seems to me like it should, but this spreadsheet seems to make it seem otherwise. Is there maybe a better way to estimate relative air flow at a given rpm?

Insight on any of these issues or others is appreciated.

By the way, all turbo calcs were done using equations found in “Turbochargers”
by Hugh MacInnes which I highly recommend:

 

neat little tool.

I have this page bookmarked also......

http://www.turbofast.com.au/tfcalc.html

 

Hmmm...that's a good link. Seems to check with mine almost exactly. The link however seems to use about 10.9 hp per #/min of flow. Maybe by using only 10 flat I was cutting myself a little short. Any opinions?

The airflow vs rpm that the link uses is the same as I used. That's a linear relationship between rpm and flow. Like I said before, this sounds correct but, after doing the math, I have to question it. If you open my spreadsheet for example, I calculated hp and tq at 3 different engine speeds. 100, 50 and 20% of max rpm and according to these conventions the torque is constant. Infact, a linear relationship between flow and rpm means a linear relationship between hp and rpm due to the 10 or 10.9 hp per #/min of flow assumption.

If this is true, which unfourtunatly it isn't , torque wouldn't change at all over the entire rpm range. And also, the hp 'curve' on a dyno sheet wouldn't be a curve, it would be a straight line.

WTF???

 

10 hp per # sounds low to me. I thought if you used an intercooler it was closer to 20 hp per #.

 

I'll take a better look at your actual equations and results later (primarily because the hot rod article leaves a lot to be desired in their descriptions, but if you ignored their numbers and used mcinnes' numbers you should be OK), but initial impressions:

- real world, 1lb/min is = to a little more then 10hp, but it's a reasonable, conservative estimate. This would only be correct if the engine is properly fueled and not detonation limited. Intercooling has NO effect on this.
- torque and VE are always proportionate. If you always have the same VE you'll always have the same torque at every rpm. I have no idea where people get the "standards" of 85% ve for 2 valve engines, 95% for 4 valve... do some math and you'll find that there are a lot of 2 valve engines getting in the low 100% VE range and a lot of 4 valve engines not touching that. It's more up to the combination then it is the engine design (where the max VE happens is up to the combination and design).
- I doubt that you'll ever see over 100% VE on a turbocharged engine without taking into account the actual VE. The restriction of the turbine in the exhaust stream just about kills any chance of scavenging and with it a killer VE.
- I don’t get the 3 points thing. If anything most engines would tend to have their highest VE on the middle point and drop off on both sides…

 

That makes so much sense to me now that you say it. I'll try some things to get that incorporated into the excel file.

The three points were for use like on a compressor map like this:



I found this method of turbo selection here , does this guy know what he's talking about?

 

I can't figure out what you're doing with the temp/efficiency stuff on the turbo size table. I found your hidden y term (line 24), but I have no idea why you're subtracting 1 from it and in the rest of your stuff you're using *R in your equation but not converting back to *F and the numbers are all wrong.

What you need to do is:
- take your inlet temp and convert it to *R
- use that to compute your ideal temp in *R (inlet temp * delta P ^.283)
- subtract the inlet temp from that to get the ideal delta T
- divide the ideal delta T by the adiabatic efficiency and that will give you the actual delta T
- convert back to *F and add that to your inlet temp

(did I just make this clear or muddy? What does McInnes say? I can't find my copy but I seem to remember him simplifying this some)

 

heh, I started reading that… not sure that he's making much sense (I won't say any of this is right 'till I see someone calculate a density ratio which in a quick scan of the page I don't see, the rd in your spreadsheet is wrong). I'll take a better look at it later (hungry, cranky, food now…), it's got one of the funky graphs that you're telling me is showing 3 points that I only can find 2 on, so maybe I can figure out what that's about if I read it…

 

http://www.gnttype.org/techarea/turbo/turbocalcs.xls

 

That page really has a pretty good description. I'm not sure that I like the discussion about the surge line, but otherwise it's very good. I'm following what you were getting at with the 20% at 0 boost now and that is probably a very safe way of estimating it (there's no way most single turbo setups will surge in that range, though I think some twins might come close).

 

WOW!

I'm going to have to go digging through his notes, but that is about the most complete setup that I've seen. It agrees with what I did myself (I haven't bothered to "publish" it anyplace since I haven't taken the time to label/comment any of it so no one would be able to make much sense of it unless they already knew how to do it themselves anyway).

2 things that I didn't bother working out is the compressor inlet vacuum (mostly because I firmly believe that if you have any real vacuum there you've designed something incorrectly) and the exhaust properties/turbine power recovery (haven't been able to find the compressor head/polytropic stuff that I've looked for anywhere). I love then airflow/BSFC/HP section… I hadn't bothered to put those together and had to look up numbers on 3 different sheets to get that info.

Too bad he's got the spreadsheet locked, I'd love to add a few things to it to make it more useful: Mach conversions in the pipe velocity section, possibly convert it to calculate a steps in a range of pressure ratios and then graph the output (mine does that, it basically lets you enter a known combination and then you can just look up how hard you'll have to push it to get the results you're looking for, makes it easy to answer "how do I run___ with this car" questions).

So who is this J.D. Estill guy and how to you get a hold of him?

 

As you see it is on "The Turbo Regal Web Site".

I got this from a turbo/supercharger topic on the Corvette Forum and I don't know anything about the author of the spreadsheet.

You could probably find more information about him at the http://www.gnttype.org forum.