107 | Analysing Turbo Performance

During the webinar we had a question about turbine flow maps and why the mass flow values are lower than the inlet mass flow values. I couldn't at the time offer a simple and accurate answer however I will do so here as it is a good question that others may also wonder about.

The turbine flow is corrected to account for pressure and exhaust gas temperature at the turbine inlet. This is why the mass flow values are smaller than the inlet mass flow values.

I've attached the equation used. Thanks to Lith for sending me this.

This is a bit off-topic, but related to something I remember you mentioning in the webinar. It's been a few days since I watched it, so forgive me if I butcher your delivery. I believe you mentioned that having a large turbine housing (more suited to drag racing for example), providing less than 1:1 backpressure allows for extra boost to be utilized (all other knock considerations permitting).

I'm trying to decide if my .89 a/r turbine housing is really necessary or if a .73 would be better for the 58mm turbo on my 2.0L in attempting to make the widest powerband possible for circuit track days with street cruising and the occasional drag race. If most guys seem to give up on this turbo around 300-320kpa, I'm pretty sure I can get sufficient results at that level of boost with the .73 housing (probably between the 1:1-1.5:1 backpressure range).

I'm wondering if the .89 housing (probably 1:1 backpressure or maybe less) would likely achieve more than a couple hp per pound of boost if I wanted to run beyond the common 320kpa or if it's likely to fall off after 320kpa with either the .89 or the .73.

My explanation in the webinar was that we find that as the back pressure drops below boost pressure, the engine starts to respond more like a naturally aspirated engine. in particular we tend to find that the engine will tolerate more valve overlap which in turn allows the use of more aggressive cam profiles. Saying the engine responds more like an N/A engine probably isn't the most intensely scientific way of describing what happens, however I find most people can understand that as a concept.

My own experience with changing different A/R turbine housings is that typically the difference to back pressure isn't massive. To get a really large decrease in back pressure you're likely to need a larger turbine wheel as well. I like to consider the turbine wheel as a coarse adjustment to turbine flow and the A/R of the housing as a fine adjustment, hence you're working within finite limits when you're changing the housing on its own.

Without knowing much about your turbo its hard to offer much solid advice. You may find that irrespective of the turbine flow, the compressor may simply be inefficient beyond 320 kPa and hence you're not going to see much in the way of power increases as you push the boost further. All things being equal, moving to a larger A/R housing will result in less back pressure and slower boost response. Since the turbine is providing less restriction to the engine, you are also likely to notice a modest increase in power at the same boost level.

First post, so hopefully not a dumb question.

I have been thinking about the adiabatic efficiency measure and in particular how it took quite some time for the measure to settle down near the published compressor map island efficiency value. If I remember the webinar correctly, this calculation was based on pre and post compressor measurements so I can't think why it would take so long for the measure to stabilise (I could if there was a large volume of air to be heated and therefore the average took a long time). Based on google it looks like the formula is basically change in temp pre / post (which must be a function of a defined time period).... so does the length of measurement matter (ie transients vs steady state)?

And given engines in real life see a more transient load, how important is that last 2% in compressor efficiency (assuming staying away from the choke line for the multiple reasons you discussed).

@CamB the adiabatic efficiency equation takes into account inlet and outlet temperature as well as pressure ratio. My 'on engine' testing is quite different to how turbo manufacturers achieve their compressor maps on a test rig and there are a number of factors that would affect the transient outlet temperature including the distance of the sensor from the compressor discharge, heat transfer/dissipation through the charge plumbing, and response time of the sensor. The sensor to some degree is probably also affected by transfer of heat from the cooler charge pipe into the sensing element so there's a lot affecting the transient temperature. This isn't relating to it taking time to heat the air exiting the compressor but rather the time it takes for the temperature in the charge pipe to reach an equilibrium.

While the engine in the real world will experience transient load, if the compressor efficiency was 2% better we would still see a small reduction in outlet temperature regardless whether we are experiencing a transient condition or steady state operation.

Makes sense thanks. Plus as you point out, better is better.