Thread: 4BD1T Turbo Sizing and Performance Prediction.

Again you are not talking about our low tech slow revving diesels, you are talking about a engine built to turn 5-7000 rpm and build big horsepower. Of course when you get to those sort of engine speeds tuned length make a difference. With a slow revving diesel that you want to build torque as early as possible, YOU WILL NOT BEAT A LOG MANIFOLD, and ceramic coat it for best performance.


Edit, in addition to that these high tech diesels have multiple injection pulses per revolution.

Originally Posted by Bush65

Possibly the best book written on turbo chargers is Turbocharging the Internal Combustion Engine by Watson and Janota. They devote chapters to Constant Pressure Turbocharging, Pulse Turbocharging, and Pulse Converters, as well as most other issues on the subject.

The pic below is part of one page at the beginning of chapter 8, which disputes some of what has been said here in a recent post.

I don't expect AM to see much gain or loss from his equal length manifold. However he could have seen a significant gain had he used a pulse converter (I'm talking about the type pictured on the bottom of the page from that book, as it was an early example).

Whether the gain for the amount of work is worthwhile is a separate issue.

I have added a pic of the factory race version of the Mazda Skyactiv diesel. This is an exceptionally low compression ratio, 2.2 litre diesel, and modified to 2 litre for Le Mans. In both displacements it produces over 400 HP. The exhaust manifold is integral with the head so not having an equal length runner manifold has not prevented good performance.

I recall seeing the Audi diesel used at Le Mans and it had a nest of snakes exhaust manifold, so they must have thought it worthwhile.

The second picture you posted is basically a log manifold integral to the head.

Originally Posted by two up

Again you are not talking about our low tech slow revving diesels, you are talking about a engine built to turn 5-7000 rpm and build big horsepower. Of course when you get to those sort of engine speeds tuned length make a difference. With a slow revving diesel that you want to build torque as early as possible, YOU WILL NOT BEAT A LOG MANIFOLD, and ceramic coat it for best performance.


Edit, in addition to that these high tech diesels have multiple injection pulses per revolution.

No John isn't talking about 5-7000rpm engines with pulse converters. It is used on engines much slower than ours.

We are also not talking about tuned length. Tuned length only works well for intake manifolds in non turbo engines. In turbocharged engines the constantly changing gas temperature influences the sonic velocity too much for tuned length to work.

The multiple injection phases in modern diesels don't change exhaust design. Unless you're trying to burn clean a CAT or DPF.

Almost all diesel R&D comes from the companies that build engines to transport large loads over long distances, not high revving engines. Ships, locomotives and long haul trucks. What is considered modern technology for small diesel vehicles, such as unit and common rail fuel injection, was in use for large diesels long before it was employed it to meet emission targets for small cars.

Pulse turbocharging and pulse converters (for engines that can't be divided into multiples of three cylinders) came from the R&D for large diesels and was nothing to do with ricer petrol engines.

Log manifolds are so widely used because they are cheap to manufacture in large numbers. When modifying a turbo diesel such as we use for more performance, it is easy to simply add fuel and air to reach the mechanical limits of the mechanical parts, without building custom exhaust manifolds to extract the last bit of performance. However if that last bit of performance is required, then a log manifold will not provide the best means to get it.

Edit, I don't see the connection with multiple injection events per cycle. BTW multiple injection events per cycle will never, ever be able to produce as much power from a diesel engine that can be achieved with a single injection event.

Am I right in saying one of the biggest factors in sizing a turbo to the 4bd1 is the compressor wheel size? A good fit being 60mm on the big side (exducer?)?

Scouring ebay for the cheapest turbo ever, I found this one with a 60mm compressor. What else am I looking for in these specs? I see there is no turbine A/R... is that weird? An important factor?

Trim is another factor that I've been "learning". In some ways is efficiency? Or at least a measure of air flow? It's not listed, but for this one I have calculated it as 55.

How am I doing?

T2 T25 T28 Turbo Nissan 200sx 180sx S13 S14 SR20 SR20DET CA18DET Turbocharger | eBay

Originally Posted by Judo

Am I right in saying one of the biggest factors in sizing a turbo to the 4bd1 is the compressor wheel size? A good fit being 60mm on the big side (exducer?)?

Scouring ebay for the cheapest turbo ever, I found this one with a 60mm compressor. What else am I looking for in these specs? I see there is no turbine A/R... is that weird? An important factor?

Trim is another factor that I've been "learning". In some ways is efficiency? Or at least a measure of air flow? It's not listed, but for this one I have calculated it as 55.

How am I doing?

T2 T25 T28 Turbo Nissan 200sx 180sx S13 S14 SR20 SR20DET CA18DET Turbocharger | eBay

It's the inducer (intake) you should be looking at. 40-45mm generally suits our engines.

That one will work (assuming the specs are correct). But honestly I'd rather take a TD04HL turbo. Their maps suit our engines better.

If the specs are correct it's a 55 trim 60mm wheel. The 60 trim 60mm wheel I ran surged with the boost I was asking at lower rpm.

Originally Posted by Judo

Am I right in saying one of the biggest factors in sizing a turbo to the 4bd1 is the compressor wheel size? A good fit being 60mm on the big side (exducer?)?

Scouring ebay for the cheapest turbo ever, I found this one with a 60mm compressor. What else am I looking for in these specs? I see there is no turbine A/R... is that weird? An important factor?

Trim is another factor that I've been "learning". In some ways is efficiency? Or at least a measure of air flow? It's not listed, but for this one I have calculated it as 55.

How am I doing?

T2 T25 T28 Turbo Nissan 200sx 180sx S13 S14 SR20 SR20DET CA18DET Turbocharger | eBay

Compressor should be sized for your desired performance. With a diesel, from desired performance the mass flow of fuel can be calculated. From there the mass flow of air is calculated to burn the fuel (minimum 18 x fuel, but better at 22x).

Now the volumetric air flow is fixed by the engine displacement, rpm and VE. So the only way to get an increased mass flow of air is to increase its density. A good way to increase the air density is to pressurise it using a compressor, but that involves adding heat, which reduces density, because of unavoidable laws of physics, and because of the efficiency of the compressor. Intercooling will reduce the temperature and increase density, thus reducing the required boost pressure for a required density.

So to answer your 2 questions. Yes, and yes. With compressor sizing you need to choose one that has a map that is a good fit for the required air flow and pressure ratio, based upon the required performance at a few rpm points that include (but not necessarily limited to) when boost starts, max torque, max power.

Compressor exducer (diameter of outlet of impeller) greatly influences the boost pressure. Ideally the air leaves the impeller at a velocity close to the tip speed (a function of impeller rpm and tip diameter) and the dynamic pressure of this air is a function of velocity squared. The dynamic pressure is converted to static pressure by the (edit) diffuser (end edit) in the compressor cover.

Compressor inducer (diameter of inlet of impeller) greatly influences the air flow. Larger inducer gives larger flow. The flow is limited by the area (a function of diameter squared) and the sonic velocity in the air.

So here you have seen the effect of exducer dia squared (boost pressure) and inducer dia squared (flow).

Trim is the ratio of these diameter squared so you should see some significance. Also comparing two impellers with the same exducer (which is the size that Garrett give in their turbo designations) the choice of trim will influence the efficiency - the one with a longer flow path along the impeller vanes will be more efficient, because it better at accelerating the air from the inlet velocity to the velocity of the impeller tip.

BTW Garrett give compressor exducer dia in their designation, but many other manufactures give inducer dia - presumably they see flow as the important criteria.

Regarding turbine A/R (area / radius), it is an important factor. Again Garrett give turbine A/R, but other manufactures don't, and instead give an area (usually in square cm), or a "phi" value. These are all valid indication of how much exhaust gas the turbine can handle.

Because the temperature of diesel exhaust gas needs to be much lower compared to a petrol engine, it doesn't have as much energy for the turbine to convert into mechanical energy to drive the compressor. Particularly when we want boost at low engine rpm (much lower than a petrol engine) the mass flow of exhaust gas is low. So we need a smaller turbine. As the engine accelerates the flow increases and when the turbine capacity (area, or phi) is exceeded, we need a waste gate to bypass the excess.

Without chasing down details, it is likely (because it is for a petrol engine) that the turbine housing on that particular turbo is larger than ideal. But turbine housings can be changed relatively easily.

Originally Posted by Dougal

If the specs are correct it's a 55 trim 60mm wheel. The 60 trim 60mm wheel I ran surged with the boost I was asking at lower rpm.

Next baby step - on this point alone, is it true that:

Surge = too much air flow.
Comp inducer is the key factor in air flow (as per John's post).
Therefore if the comp exducer is constant at 60mm and a 60 trim is surging, then a 55 trim (smaller inducer) should have less or no surge?