Volvo Project - Part 2 [ February 7th, 2011 ] By: Mark Ozimek Posted in » Ramblings
I mentioned in part 1 that I have a hard time making up my mind. This is the story of how I came to decide what turbo should get bolted up to the engine to get me to where I want to go. Be forewarned: A lot of text lies ahead. I’ll do my best to be interesting as I tell the (not so) enthralling tale.
As a point of reference in all this, for those who are unfamiliar with the Volvo powertrain, the stock shortblock seems to be good for around 600hp without sleeving, assuming the engine tune is good and heat is managed properly. Beyond that, the cylinder liners have a tendency to crack where they touch the next cylinder. The 5 speed transmission, M56H, is reliable for around that much as well, and can handle more, although gear and bearing life is rapidly declining at that point.
Originally, I was aiming for around 350whp, maybe a bit more, with a 56 trim Garrett GT2871R tucked away behind the engine. Let’s take a look at how the engine matches up with the compressor map. I made some very basic and incorrect assumptions that will get me into a ballpark estimation, such as the pressure ratio across the turbine being equal to the pressure ratio across the compressor. That will give a rough feel for where the boost threshold lies.
This is at 21psi, with a 7000rpm rev limit. Because I am looking to make this last a reasonably long time, I am choosing to keep the shaft speed around 90% of the maximum listed on the compressor chart. For the GT2871R, this is a whopping 120,000rpm! This allows for some special circumstances, like driving up mountains, to avoid overspeeding the turbo to hit the higher PR needed to get target boost in thinner air.
Anyway, onto the actual graph. As you can see, this turbo looks pretty well matched to the engine I want to build, although it is just a bit on the small side for peak power. The spool-up is based on the 0.64 A/R turbine housing flow curve that Garrett provides. Volvo uses a T3 flanged manifold, so I would get this turbo with the T3 based 0.63 A/R turbine housing, but that shouldn’t noticeably change spool.
That is just about enough airflow for about 400bhp without pushing the turbo too hard, or around 340whp. Being a FWD car, that seemed pretty reasonable figure. More would only really be usable at very illegal speeds, or on a pretty high speed track. The real nice thing about the GT2871R was that it should be making as much boost as I wanted by around 3000rpm, which is perfect for the highway, where the engine sits at 3000rpm as the car cruises at 75mph in 5th gear. Stepping up to a GT3071R or GT3076R will bring the boost up to 3500-3750rpm, which may be a bit too late for my tastes, despite the possibility of a bit more power and a cooler running engine from less exhaust restriction on a small turbine wheel.
I thought I had my turbo picked out, and had everything picked out to support it; ATP ultimate internal wastegate, the actuator, an adapter flange, the hose kit needed to get all the fluids to and away from it, the whole nine yards.
Fast forward a few months, and Garrett announces the GTX3582R, 3076R and 3071R. With a redesigned compressor wheel, they give about a 20% boost in max airflow from each turbo over the GT turbos they replace. Curiously enough, they switched from 12 split blades to 11 equal height. That will certainly affect how the compressor wheel performs. Plus they added “extended tips”, which basically just makes the compressor wheel bigger than its advertised exducer size.
Older “GT” compressor wheels look like this:
Newer GTX:
The basic sizes of the wheels remained about the same, and overall efficiency didn’t change noticeably. The general operating window got pushed to higher PR and more flow, including shifting the surge line up. By by pushing the compressor map to the right with the same turbine wheel, the compressor will be operating in a slightly less efficient spot during spool-up. I suspect this will push the boost threshold up in the RPM range a bit, as there will be more energy required from the turbine to compress the same amount of air to the same PR.
Despite previously ruling it out because of the spool time, the GTX3071R seemed like more viable alternative. It suddenly offered a much higher power potential without a significant impact on spool from before. Despite being “slow” compared to the 2871R, I reasoned that having boost by 3500-3750rpm could be doable for a DD. That still left me with about half of my total RPM range in boost, which is far from being a spiky peak hp dyno monster.
Not long after that, I found out about BorgWarner’s EFR line. There were a couple things that I really liked about what BW did with them. First, they made a really light turbine wheel, and kept the size up. This improves the turbine efficiency, and increases the amount of torque the exhaust gas should be exerting on the turbo shaft. This, along with the reduced rotating mass compared to the typical Inconel turbine wheel, should greatly improve transient response, and reduce backpressure a lot while keeping a configuration that still allows a respectable boost threshold.
In playing around with Matchbot, it seems that the EFR7064 will spool around 2750-3000rpm, and the 7670 will spool around 3250-3500rpm. As far as turbo performance goes, the 7064 stacks up pretty well against the GT2871R; similar boost threshold, potentially faster transient response, and can supply a few extra lb/min of airflow at the top end. The Garrett is better than the BW at lower pressure ratios. The most pressure I want to run on the GT2871R is about 21psi, from what we saw on the chart before. The improved performance of the 7064 at higher PR and higher flow means that I could run about 25psi and get a reasonable improvement in power without compromising the spool.
In the end though, I ended up settling on the EFR 7670. Here are the operating points found through the matchbot program, targeting a peak boost of 30psi, the points are at 2750, 3000, 3250, 3500, 6000 and 8000rpm. As you can see if you can squint hard (or right click and open the image to see the original size), it can make 30psi by 3500rpm and hold it to 8000rpm without overspinning the turbo:
I decided that having full boost by 3500rpm, going through the peak efficiency islands of the compressor wheel, and a potential for 500+whp was a good compromise, despite being more power than I should really be trying to push out of the block, and even more than I should be trying to put down to the front wheels of a street car. Logic be damned, I’m gonna do what I want! Plus, the EFR series has the distinct advantage of having a built-in recirculating BOV, and a high-flow IWG with an actuator that comes with the turbo. Those two things save enough money to make the higher cost of the EFR worthwhile.
So, one step of the project out of the way! I know what turbo I’m going with now. It’s time to make the rest of the engine support my goals. I’ll save that for part 3, since this is already a tl;dr post.
