Turbocharger compressor efficiency

[@lamazver]

I just love that you took my favorite engine in equation. From what I've seen on multiple engines of these is that we got 400hp@1bar of boost with s257 and S363sxe turbos. In one instance with insane engine there was 485hp@1bar and 8400rpm I would love that you explain to population density ratio and not pressure ratios, since pressure ratio isn't exactly showing true numbers innterms of airflow and power because it misses density we get provided to the engine by turbo and charge cooling setups...

[@TurboDirectSA]

Density is not important in a rudimentary calculation, as you would then need to take into account temps, humidity, pressures, air cleanliness and alot more - waste of time for a basic turbo matching exercise.

[@802Garage]

Funny I just left a comment on "Compressor maps in a little more detail" stating this was one of my remaining points of confusion. How to plot approximate points on a compressor map when picking a turbo for a specific engine and power goal. Then I find this video I haven't seen. You basically answered my questions in the other comment point by point. At least that means I was asking the right questions! It's done just as I expected. Guess I wish I could be lazier with the math, but my quick estimates tend to be pretty good. There are still some finer points which confuse me, such as how to know if an engine and turbo combination will be able to hit target boost by a certain RPM, but I'm getting there. Thanks as always. Nice paper weights!

[@TurboDirectSA]

Glad you enjoyed the video - :-)

[@thysellis3677]

Great Video, thanks Chris, can you do a similar video on exhaust housings,this is still a grey area for me.

[@TurboDirectSA]

Will do - in a future video coming soon

[@kevin_delaney]

Wow, thorough video, even if it's high level. Makes me curious what a full in-depth video is like 😆😈 it seems like a lot to ask, but I'm super curious haha.

[@TurboDirectSA]

Its a very broad and in depth topic and based on opinions and questions/interactions the topic is literally endless.

[@ShredCulture]

Great video. One thing that was not covered is the temperature and it's effect towards turbo flow requirement. By multiplying the pressure ratio by the mass flow N/A, you are assuming that the air temperature after the turbo or intercooler is at standard temperature 70 F (aka no change in temp). This would be fine if the intercooler was 100% efficient. However most intercoolers are more in the 75-85% range. This means that the temp going into the engine would be higher therefore the density ratio will never match the pressure ratio. In this case if operating at 3.0 pressure ratio, you would have to multiply the n/a mass flow by the Density Ratio and not the Pressure Ratio since the temperature change needs accounted for because density ratio X n/a mass flow is what the turbo needs to flow at that pressure ratio and therefore a more accurate representation of how efficient the compressor will be.

[@TurboDirectSA]

Thanks for the comment - 100% hear you, i kept this easy to understand, but there are other factors that also influence the maps and plots, but this formula will get you as close as dammit to the final result - and for most will act as a guide to help them choose a suitable compressor match.

[@lamazver]

Thats what I love about match bot you get all these calculations provided. Its awesome. I would definitiely love that borg warner make a hybrid version for compound turbo setups.

[@Christdeliverme]

Chris as always great videos. Request: In depth explanation of turbine flow limits # per HP support etc. Using Garrett's published graphs.

[@TurboDirectSA]

Thanks buddy - will do this in a future video for you.

[@Christdeliverme]

@@TurboDirectSA

[@TurboDirectSA]

@@Christdeliverme Ill share some experiences with you and what works and doesnt work and why - with custom built combinations

[@Drunken_Hamster]

So a couple things I noticed; 1. This assumes that you CAN reach 2bar gauge pressure by 5000rpm on such a small engine with such a large turbo, which, if that IS the case, then you would either want to use boost control to limit the boost at that RPM on the Borg so you don't surge, OR you would actually want to have the Garrett, instead, for the same reason. And 2, the Garrett will technically be more efficient between 6000 and 8000 RPM on average than the Borg. In contrast, the Borg will only be maximally efficient at 8000 RPM, which is fine when you're at the very end of the run and not looking to change gears anymore, but perhaps not as ideal as you're going up through the gears dropping from 8000 to 6000 and then climbing as you accelerate. Suppose the overall goal of anything is peak average efficiency within a band, and the goal of this specific car/build is peak power/efficiency at the top 1/4 of the rev range. Wouldn't you want something ideal BETWEEN 6000 and 8000 rather than just AT 8000? I could see the latter being the case if you were racing with a CVT (assuming you could make a CVT withstand the torque) but otherwise, if multiple set gear ratios are involved it seems technically ideal to have a range of average peak efficiency than just max efficiency at one RPM setting.

[@joshualamoureux2520]

How do you compensate that different turbos will flow different volume of air at the same pressure ratio?

[@TurboDirectSA]

What do you mean compensate? No two turbochargers are the same in terms of flow vs P2C and rotational Speed.

[@fernandohood5542]

Can you determine VE from a dyno run? How did you determine bfsc?

[@TurboDirectSA]

You probably can, but the VE will not be accurate, there are always losses in the drive train, and outside conditions that will affect the calculation that the dyno does to provide you with the power and torque. VE is usually calculated from the engines specifications. The manufacturer will provide the VE, or you can use the Power that the engine makes at the RPM it produces max power and its CID to calculate the VE. BSFC is mentioned in the video. Keep in mind that the calculations provided in the video are NOT in-depth, these are used to provide you with a very good idea (within 5%) of the actual value if you take into account humidity, Baro, Density Altitude, intercooler efficiency, pressure drop and and and --- so that you can identify and choose the correct turbochargers compressor for your application. Its NOT designed to split hairs and calculate the very last 0.00001 decimal - if you get my drift.... Hope that helps -

[@stanislavaleshin1202]

Once again - great video! Three questions: 1) you've mentioned that 10lbs/min for 100 HP it's a quite conservative assumption, so - for modern tsi engine what proportion is more suitable? 10lbs for 110HP? or more? 2) honestly, I haven't understood - why you chose borg wagner as the best option, spool became later and the same efficiency in comparison with Garret. I think that both these turbo are not the best option, they are way too big for 2.0L engine. Could you explain your choice? 3) One question last - how to match right turbo A/R and turbo size for exact engine and power goals, is it possible to explain this point easy? :)

[@TurboDirectSA]

Stanislav each engine is different, it all depends on the engines VE. Any change made to the engine will change its VE, which is a key value in the formula. 2) Nowhere did we speak of spool -- EFR is designed on another level compared to Garrett in terms of bearing design, longevity, attention to detail, comprehensive simulation analysis - search for the technical doc on the EFR range on BWA web page for more details - you will be surprised. Both these turbos are 100% perfect for this engine in my video - AS A 1KM CAR ....... there are a miriad of other options for other applications like street, drift, drag etc.... It is complex in choosing (based on calculations) the best AR turbine housing, you need to take into account shaft power output, flow, EGT, Fuel type etc..... it is obviously possible, but very complex.

[@stanislavaleshin1202]

@@TurboDirectSA Chris, where could I find this video (AS A 1KM CAR)?

[@TurboDirectSA]

@@stanislavaleshin1202 The very same video you commenting on.

[@chippyjohn1]

He explained the reason for choosing these turbos; for absolute top end performance. If you were going to spend more time lower in the rev range you would choose a smaller compressor, but the example was for the most efficient compression in the top of the RPM range, hence there was much map left over to the right of the peak rpm.

[@exploranator]

The skinnier you can get away with making your compressor and turbine blades, the greater will be your efficiency for an overall turbo size. The gamma-titanium turbines are FAR lighter but they come at an efficiency penalty for steady-state power production. I believe MFT's (mixed-flow turbines) may be a way to claw back some of that lost efficiency. Until then, I still prefer the idea of gamma-titanium turbines anyway, as absolutely minimizing/eliminating turbo lag is a fascination of mine, and I am not dealing with big rigs or marine engines that sit at close to the same engine state for seconds, minutes, hours, and days, so efficiency is far more important than lag elimination. I also await mixed axial/radial multiple-shaft turbos, too.

[@TurboDirectSA]

The Gamma-TI blade designs are mixed flow.

[@exploranator]

​@@TurboDirectSA Well well well. Borg Warner is intent on taking over the world aren't they? I am curious if an even more aggressive flow mix will be done next, to more closely approach the axial turbine while still preserving the energy of the radial flow. The direction of the radial flow's rotation can impart a similar effect to the flow direction as do stators mounted before turbine stages in an axial turbine engine.

[@daviddroescher]

Questions Do you ship to the states, and do you have your .71 v band v band special tongue Garrett fitted for a G25-550/660 For sequential , youd divide the cfm by 2 and do a map for each bank, correct? For compound? A feeds B. B feeds the motor (B =manifold/small /high pressure charger) ?run the numbers for A. Then using results from A in PsiA as the inflow for B and run the numbers agen for B/B's? I understand it is way more complicated than that, but this is my best guess. ?how to compensate for work induced thermal expansion going into B (map is in "corrected" lbs/min) aka how to un correct the flow rate planed phase 3 setup for 175mph on 160hp worth of fuel looks like this G25-550/.72ewg (B1 back 3cyl/2.9l) g25-550/.72 vband ewg reverse rotation ( B2 front 3cyl 2.9l) fed by a s362sxe/14(A 5.9l) peek whp on dyno was @1976. So low rpm 130-150mph Series-parallel 1xA into 2xB Will DeltaT help to confurm my location in the efficiency map. Eg -7°f air filter, 257°f compressor discharg and -11°f intake manifold@ 15psig/27psia(7k' above sea levle, look up hypoxia drag strip ) 1550 rpm /70mph Ambant -15°f Delta T compressor =+264°f Delta T IC = -2662°f (4° above ambant)

[@TurboDirectSA]

No we dont have any housings for the Gseries turbos as yet.... coming soon. Sequential - controversy, mixed understanding .... sequential="in sequence" how in sequence? Everyones idea is different - without knowing what your idea is, its impossible to comment. What size turbos, what AR housings, location, runner length, ID dimensions? Very complicated to setup depending on what you want to achieve. Compound is actually quite simple - BUT it all depends on what the intended requirement is once again. Compound turbo setups are designed for massive power outside of a said turbochargers ability to be able to provide sufficient HP and the engines ability to spool it up. Mapping is technical - shoot an email off to [e-mail] for me to look at for you.

[@zainulpatience406]

great video ad great explanations

[@chippyjohn1]

My maths is; half capacity in CC, X RPM, X air density per CC, X VE, divide 60, for grams per second (1 gram per second=1kw) Multiply by the desired PR Or ;half capacity in CC, X RPM, X air density per CC, X VE, /1000 (grams to kg), X 2.2 (KG to Pound) for 10 lb/min for 100 hp. Multiply by desired PR. 2000CC /2, X 5000, X 0.001225, X .90 (VE), /60 = 102 grams/s or 102kw at 5000rpm. X1.5PR =153 grams/s or 153KW

[@TurboDirectSA]

Thanks for the comment -- the above is a little too complicated for a compressor map outline for beginners or someone who wants a basic and quick ref point. To comment on one aspect of many - you have no BSFC formula on the turbine side, which is what drives the compressor - these forumlas become a function of the turbine effects when using the above relating to air density. Without the Shaft power and BSFC forget about anything accurate. This is why i kept the outline very low level - not to lose the guys with less understanding. For the more in-depth calculations i use Match Bot - its the most advanced and comprehensive tool there is. No need for formula's or a scientific calculator.

[@chippyjohn1]

@@TurboDirectSA You do not need a BSFC for the turbine side. It just works on mass and pressure ratio. It is more simple than the compressor map. That is my finding. I had to teach myself about maps and A/R as there is no help for my project. Garrett refused to help me. I have to evaluate 5 stage compounding for my project, not the usual car project. Do enjoy your videos, wish they were more in depth though. You wouldn't have any information on what aluminium alloy Garrett and BW use for their milled compressor wheels would you?

[@TurboDirectSA]

@@chippyjohn1 Nope - the majority of the turbine side works on the fuel, and BSFC - id strongly suggest you play with match bot - without BSFC forget about any information and/or results from the turbine side - which has a direct effect on the compressor performance, based largely on the turbine wheel shaft HP output. Making a small change to the AFR, EGT and type of fuel will move your turbine point plots with massive conflicts. Some will land up on small AR housing options and others on much larger turbine AR options - try it and you'll see.

[@FWD_Problems]

have you found a compressor map for the stock is38? I cant seem to find one.

[@TurboDirectSA]

No - not really required, the turbo is not a very powerful unit at all. We have never needed a map as its not a really attractive upgrade for many retro fit applications.

[@FWD_Problems]

@@TurboDirectSA I more wanted to see where my car sits. you know, try out your math and learn something. :P

[@TurboDirectSA]

@@FWD_Problems I understand - Ill try source this from IHI for you.

[@mattmartin598]

Love your stuff. I'm still having difficulty with my build. I know you said in another video that you think the 8474 is a great combination with an EJ257. I'm discussing my build in a tuning discord group and there's a guy there who says it's the worst turbo he's had on the engine, 0.83 seconds lag between shifts. 2.5, fully built. I know there are many variables, but I'm trying to understand how he was in love with the 8374 but hates the 8474. I'll be destroked to 2.34, closed deck, 14mm studs, cams, springs, etc.. Shooting from the hip I thought the 8474 would be great, but this guy is making me doubt and now I'm trying to get more into the science of matching the turbo to my build. I'm hoping you might be able to point me in the right direction. Thanks!

[@TurboDirectSA]

I battle to believe that the time between gear changes is so poor with the 8474 compared to the 8374 - the turbine and housing stage is identical --- i believe there might be other problems that are causing this lag, its not going to be from the turbocharger itself. What HP target are you planning on reaching?

[@mattmartin598]

@@TurboDirectSA I'm aiming at 800whp without changing intake manifold and with a 1.05 A/R EWG. I'll be tuning on a substantially low reading mustang dyno, probably in part due to its altitude over 2700ft, so it's a tall order. I thought the exact same thing about the shift lag difference, and have been trying to dissect his setup to pinpoint why. He's running emtron, as I will be as well, so he's got plenty of graphs showing lots of details that go above my head. 0.83 sec to get back into it, shifting at ~6700 rpm, turbo speed (seems low?) at around 95k. Something has to be off. I can see boost threshold increasing, but I can't find a reason for the lag to increase. And a high revving 2.34 should be plenty of juice to keep the 8474 in a happy range from my research. Thank you for your time 🙏

[@TurboDirectSA]

@@mattmartin598 The AR1.05 is too big. The biggest housing you should use is the T4 AR92 - build a proper equal length runner to the up-pipe. What gearbox you running? RPM drop between gear changes should not fall below 5000RPM when shifting at 6700rpm - so you will definitely be in boost after changes. The 95K turbo speed indicates 1.5-1.6bar boost - --

[@mattmartin598]

@@TurboDirectSA 1.05 is the smallest they offer with external wastegate option.

[@TurboDirectSA]

@@mattmartin598 Correct, it will be laggy. You will need to shut off the swing valve or perhaps just use it - it measures 42mm diameter, and will respond faster than any external gate will.

[@dennisclanton5342]

Thank you

[@ianjohnson4170]

@TurboDirect S.A - K20A2 built motor, 11.0:1 Compression spring retainer motor. We are planning to run 600whp at 9200rpm, im trying to find a turbo that will work for a mixture of low boost Autox (400whp) and Street (600whp). Looking for something that will spool before 4krpm that can meet both conditions. Is the G30-770 one of these? is there anyway you could help me with a plot?

[@TurboDirectSA]

Ian - i find a few issues with your requirements. 1) Are you requesting that a turbo make the peak HP of 600WHP at 9200RPM? If so, this turbo will be large and will never spool around 4000RPM. If NOT - then are you requesting that the turbo make its power earlier in the RPM range and lower to 600WHP at 9200RPM? - If so, the once again this will be a relatively large turbo and most likely will not spool before 4000RPM. Any turbo that spools BEFORE 4000RPM will be relatively small considering the high compression you running, and will not be as small as the EFR7163 or EFR7670, but more in the line of the EFR8474 Black Series (950HP capable) which should be able to spool up around 4000RPM full boost with that high compression, ensuring that you using a decent fuel which will allow a decent ignition advance without hurting the motor. This turbo will make the power you need, and is able to make it earlier in the RPM range, but with clever boost control, you can definitely have the 9200RPM mark still making 600WHP - provided that the engine can use the flow form the turbo and convert it to HP - G30-770 is a waste of time on this application because of the high compression and high RPM band - The power will guaranteed fall away from around 6000RPM even with high boost and controlling the boost in the high RPM area by increasing it as the engine revs - the same will happen with the G30-900 - I would suggest the GTX3582R Gen2 or G35-1050 turbo for this - either will work, but the later is much more expensive.

[@ianjohnson4170]

@TurboDirect S.A - yea we are planning to do time attack and autox at 400whp and are wanting as quick response as possible. Our street setup will be 600whp area. All on E85. Im not great with turbo selections and was looking for some advice. So a Gtx3582 gen2 will keep good response around 4k to 4500rpm and still be capable of 9200rpm without dropping off?

[@TurboDirectSA]

@@ianjohnson4170 It will yes - but what i would do is be clever with boost control to hold the power and torque curve up in the higher RPM range - if you find (gear ratio dependant) that on the tighter corners in Autox you losing response, then introduce a slight antilag setup to help you out there. Then you'll have the best of all worlds.

[@strokerz2858]

it's so funny 😂😂😂 Garrett GTX3582r gen 2 50hp by changing the catalog volume

[@TurboDirectSA]

Very disappointing indeed - the dream team that once was, is gone since Garretts' bankruptcy in 2020 - quality is not what it used to be, nor is the performance of the product according to their claims.

[@strokerz2858]

@@TurboDirectSA Maybe you can convince them to come to South Africa to work for you...🤔

[@TurboDirectSA]

@@strokerz2858 LOL - one of the best guys in the business was Trevor Cass - he has retired now. The divorce between Honeywell and Garrett years back, lost Garrett the aerospace design teams knowledge, as these were the guys who invented the GTX Gen1 range, which was a massive step ahead in the compressor design in 2010 - certain people in management is all i am told - caused this mess. Such a pity - great people in the old team, all gone.

[@daviddroescher]

BSFC how could one figure this data or is this a dyno only figure , from mesuring actual fuel consumption per each hp live? I am thinking the the data provided by Cummins 72cc fuel per 1000 events= 72cc /min @ 2000rpm 160bhp @2500 (fake news Chassie dyno said 45whp @2500 without correction ) Rear world With correction 173whp@1976RPM (peak hp and start of governors fuel cut) So Rpm ÷ 2 × single injection volume ÷ BHP=~ BSFC 1976 ÷2=988 ×.072cc=71.136cc ÷247.1=BSFC 0.28/hp Wheel hp (WHP) to crank hp(BHP): Whp÷7×10=bhp or whp+30%=bhp 247.1÷7=24.71lbs air/min whp 24.71×10= 247.1bhp.

[@TurboDirectSA]

To keep it simple as possible - BSFC is the amount of fuel burned for a specific amount of power produced. Each fuel has a different calorific value and hense a totally different AFR. So using pump fuel as opposed to ethanol (different percentage content will differ the number used) or Methanol will have different BSFC - this is why the Match BOT from BorgWarner is so good as its THOROUGH and requires the fuel you intend using to make its final calculations ... are you starting to see why Borg is better... if not you will soon enough.

[@daviddroescher]

@@TurboDirectSA yes I understand the "what" BSFC is and how valuable it is as data. However "how" to get said data(BSFC), I need help. The local (700mi drive) chasie dyno dosent have the ability to mesure fuel flow on a MECHANICALLY Injected deisel. I know it is easy just strap $5k(each [2 needed]sender $2500 to accurately mesure my 2.2ga fuel flow /hr) of lab equipment to the motor drive for a little wile and wala I've got the data to do the math to find the right turbo for my Non-standard build(2000lbtq@1500rpm).

[@TurboDirectSA]

@@daviddroescher You going to need to contact the manufacturer of the engine to get the necessary data on it, and from that calculate the BSFC and use that value - the manufacturer will have this info for you though.

[@JaySimons1]

I dont quite see how this info will help select a turbo, you didn't explain how you know it will get into boost, ever! For example even bigger turbo on that engine may never get into boost right? just placing a dot on there at 50lb/min doesn't mean it will get there. So more to the point how would I calculate what rpm I might get boost with a given motor and turbo, that would be useful for me. This video makes it look like you can put any size turbo on a 2L and put a dot on the graph of a good brand turbo, and it will get there, which I know is impossible IF the turbo is too big.

[@TurboDirectSA]

Jay from around 1:05 until 1:35 i mention that we will not be going into the low level in-depth calculations and we wouldn't be splitting hairs about every single formula and contributing factor - this is a basic outline on how compressor maps work, how to plot on the compressor map and how to calculate the plot points based on your specific engine build - its not discussing or demonstrating the turbine calculations (which is extremely complex in comparison to this video) which is what you will need to be able to plot the "response curve" and understand which AR turbine housing to use etc -- This video is as i mentioned at 1:34 "This is the basic jist of how to plot a compressor map" at 4:13 until 4:25 i also mention that we not going to try and plot low engine RPM - hense we started at 5000RPM upwards. This video is designed to help most people choose the correct turbocharger - based on its compressor map. This video certainly does not make it look like you can put any turbo on a 2L engine - do the math and you will see that for example - trying to use a GTX5533R (obviously too big a turbocharger for this 2L engine) will not even put you anywhere on the compressor map past the surge limit under 7000engine RPM -- and the same goes for trying to use the GT1241 turbocharger, which will put the plot after 4000RPM past the choke limit of the map --- If you looking for response, and EMAP, and turbine AR choice, then the Matchbot is the tool you need - this is the easiest way to plot the turbine side of the turbocharger.