Mike Shields on Closed/Open Deck

Subject: Closed deck/open deck production dates and WRX/2.5 RS engine differences.

First, a polite apology and a disclaimer.

In checking on Trey Cobb's inquiry, I found a chart that shows there were open deck WRX turbo motors in production since late 1994. I had done major research on the WRX in the winter of 1995-6 for my own amusement and reported on the incredibly detailed, original, 1993 Japanese language production information available to me at that time. The information provided was correct for the time period reviewed. No new published information since then was found until this summer when the HyperREV ver28 was available at my local Japanese book store in Torrance. Even then, I did not read it cover to cover, as reading Japanese for me is about like pulling teeth. As some of you know, my German and Chinese is much better, while it is my wife who has the Japanese down and she could care less about cars, so I only ask her when I need to.

At the time the 2.5 RS came out and when comparing North American engines with Japanese engines, the open deck and closed deck issue is was the most obvious, but not the only difference. While I see that open deck turbo motors are used, it is my honest opinion other issues remain regarding any decision to turbocharge the 2.5RS engine. It is obviously now the time to list these issues, based on conversations with Subaru staff and Prodrive staff, as I have done in the later portion of this article.

For those of you new to the list, I manage SPD Tuning Service in Redondo Beach California. Because of our business plan, we are very conservative in developing and offering performance parts. If one were to purchase a C43 Mercedes-Benz, one would simply not question the quality and reliability of the AMG modifications. One might question the price, but that is different issue. Quality is what SPD Tuning Service is aiming for. Not the cheapest, not the most trick, just the most effective solutions to real performance needs combined with complete satisfaction from every product offered. We cannot ask our customers to plunk down $4000 dollars for a turbo kit when the majority of them are making car payments and often do not have alternate transportation, unless we are absolutely sure that such a kit is going to work and work properly for a long time. Please realize that my opinions and comments are always based on this business outlook. Others are free to disagree with the opinions presented here, but not with the principles on which the opinions are based. All I ask is that "we can agree to disagree."

With this agreement in mind, I remain skeptical of major, reliable, turbo power from the 2.5 RS, at least on a reasonable budget. I have also said in my published comments at http://www.spdusa.com/ , "I would love to be proven wrong." Wrong, if only for the reason that some of us may be willing to spend a couple of thousand in engine related maintenance costs every two or thee years. As a result, ones assessment of the risks involved with turbo kits will be different. For some, a turbo kit for 2.5 engine will be very desirable, no matter the costs involved. Here is my position: I want to have a kick-ass motor I can run for not months, but for years of reliable service.

That is why, some 15 years ago, I went from my second 1.8 liter 145hp "SSS" engine straight to a 195 hp, fuel injected Nissan V-6 engine in my 1973 Datsun 510. Worked just fine. With my Impreza, I am going from a 118hp 1.8 engine to a 260hp WRX engine in the same way. As Walter used to say, "And that's the way it is." End of disclaimer! Please read on.

Because Trey asked, "Are you sure? Check HyperREV!" I started looking closely through the entire HyperREV with my Japanese speaking wife at my side and can report the following information for Impreza WRX Turbo engines as of 97.9.1. This information can be found on page 176 of HyperREV Number 28, printed in 1998. Here is what I found: Prior to 94.10.7 all engines are closed deck at 220hp (for auto trans, w/hydraulic lifters, slightly smaller turbo w/milder cams), 240hp and 250hp (STI). Note, I feel the autotrans engines are desirable as they are less likely to have been trashed and SHOULD actually run on our 92 octane at their designed horsepower and torque. Standard production "WRX" EJ20G engines changed to an open deck design at 94.10.

This is where I missed the boat, as I had only the original 1993 factory manuals. The WRX-RA, WRX-STI and STI ver II remained closed deck with a high performance cylinder head, and other goodies, up to 96.6.6. This closed deck "RA" engine is the "big power" motor that made Subaru famous in WRC (world rally championship). This RA version block, head and forged crank are still the ones to get if "going for broke" in the power department. The RA head in particular has different water jacketing, very expensive valve material and a host of other details, while the block remains the strongest EJ series made, along with its cousin, the EJ22 turbo block.

The EJ20G "WRX-RA" motor was designed for an era when Group A rally car engines were seeing 400 plus horse power in the 8000 rpm range. The rule change to 36mm and then to a 32mm restrictor on the compressor inlet reduced the power output to the point that special 1996 Prodrive road prepared WRX-RA and STI version cars were "faster in a straight line than the Group A cars". This is a quote form David Englishby of Prodrive over lunch in Banbury in 1996. The inlet restrictor, with the lower rpm, lower maximum expected power figures along with the Group A rally car rule change to permit special engine and chassis developments, independent from actual production cars, lead to the redesign in 1997 of a new "EJ20K" series engines.

The 1997 year model change over date, "96.6.6", saw a major internal and external redesign of the engine. It is now the EJ20K series with re-routed, resized and repositioned turbocharger, repositioned larger intercooler, lower compression 8.8->8.0, bigger valves, more (relative) boost, and slight cam changes. After this time, all production engines are open deck design, including the STI and RA versions. This permitted a great cost saving by changing all engines over to open deck die-cast cases while keeping the basic strength for a 300hp design limit.

One of the key developments was a ceramic coating process that was first done by a small UK company for the Prodrive assembled racing engines. A manufacturing process was developed for this piston coating, making it possible for the pistons of the K series engines to have greatly reduced friction surface with the cylinder wall. That the "K" series engines have an open deck is not the end of the story on whether to turbocharge the RS 2.5, as we shall see. In addition, From 96.6.6 a special EJ20G series at 250hp remained in production all along with hydraulic lifters, 9.0 compression and a smaller turbo. This engine was the basis of the Australian and UK Turbo models at 218hp. While not mentioned in HyperREV, they have now gone to a "K" series, with its changed layout of peripherals, retaining the 218hp specification, small turbo and higher compression. These are ideal candidates running on US gas, if a low mile, salvage engine and transmission can be located. Note that these transmissions have a 3.9 final drive compared to the 4.11 final of the RS 2.5. However, their 5th gear is shorter, giving similar revs per mile in 5th. This trans is "autobahn" gearing, much like the Audi 90, where first and second handle everything up to 45-50mph and then three more gears to 130 mph. We tend to want stop light to stop light performance and get shorter intermediate gearing.

There is obviously a ground swell of interest (if not actual cash) in finding some way of turbocharging the US engines. The fact remains that although the redesigned K series turbo motor cases have an open deck design, it does not follow automatically that the EJ25G DOHC engine used in North America is cleared for take-off. There were and remain additional issues to consider. Simply put, when first asked about turbocharging the 2.5, the open-deck issue was first on the list and made further discussion to my mind pointless. Lest we give an answer in search of a question, let's ask this question: Does the 1998 2.5 RS DOHC or 1999 SOHC engine have common design features with the open deck or closed deck turbo motors? Here is the information I have been able to assemble. It may not be the final answer, but it is what I have been told.

We know that the issues that resulted in the open-deck K series motor were answered as a result of lower demands being placed on the engine due to changes in racing rules and from actual design changes made by Subaru engineers. However, one can not assume turbocharging has been evaluated in the EJ25 engine. The following remarks were made to me by a Subaru engineering person, who should remain anonymous. They stated that the 2.5 DOHC engine was a "damn the gas mileage, give me some torque" responce to US market needs and is a "stretched" design specification engine. As a "stretched" design, it was produced with no consideration of additional mechanical/thermal stresses of high power and/or rpm. For the sake of the argument, its power output at 165hp is as near its design limit as a "K" series WRX engine is at its 280hp output.

In contrast, during the same conversation, the closed deck Japanese EJ20G turbo block and the older 91-93 American closed deck EJ22 turbo block were considered "bullet proof". This person had no personal knowledge of open deck WRX engines at the time we spoke. It was specifically mentioned to me that the thrust bearing area is known to be marginal WHEN CONSIDERING MORE POWER. Note the emphasis. Not that the engine has a design problem - that was not the point. The point was if one were to take the engine out to some big torque and horsepower numbers, the bottom end was a limiting factor. This implies that the bearing webs in the case would "walk" under the torsional loads of large power output, where at the designed 165hp this was simply not an issue. This type problem has been true for many engine designs over the years, by the way.

Second, the cylinder head and water jackets were not designed for the thermal stress of turbocharging.

Third is the thickness of the cylinder wall. This is a particularly important issue when it comes to the head gasket area in an open deck design. From the brief conversations I have had with Subaru staff it was made clear that no special attention was paid to potential stress from high output in the design process of the EJ25 DOHC engine. In contrast, the design of the EJ20G Turbo motor saw every area of design was gone over in the "WRX-RA" series engines with a mind to 400hp output. This is the most important point that one needs to know when assessing the 2.5 motor.

Just where are the limits? I was left with the impression they are lower, rather than higher than one would assume from the rally heritage of the turbo EJ series Subaru boxer engines. What else can I say? I was sitting there having this conversation about the RS 2.5 engine and feel it should be made public. It does not mean one can not supercharge or turbocharge the 1998 EJ25 engine to a nice 250hp. It just means that SPD Tuning Service is not going to develop a kit for this purpose, as I will also very seldom bet on less than a pair jacks when playing five card poker. The odds are not in my favor. Especially with the typical urge, myself included, to turn up the boost.

In the same conversation, I was told that 1999 2.5 SOHC engine has addressed the bottom end concerns. Of the two engines the DOHC 98RS and SOHC 99RS, I was told the SOHC engine is a much stronger engine. In driving the SOHC engine, it does seem to be a smoother, tighter engine on the top end, indicative of a stronger case, among other changes. The new 1999 EJ25 however, along with a 1.8 and a 2.0 version, is a family of "lean-burn" combustion chamber/engine management engines. I do not know how the this aspect will play into the high pressures and temperatures of turbocharging, nor if the valve and seat material is up to the higher local temperatures seen in turbo motors. So again, spending the big money and developing a reliable turbo setup remains an open issue for me.

On the other hand, I suspect that if we see a turbo motor, it will be a 2.X SOHC, non-intercooler engine from this new family, with the necessary engineering changes. This part of the puzzle is starting to make sense, as it is unlikely we will see the EJ20K with the extra cost of the DOHC heads and intercooler. The fact that we have only 92 octane gas would effectively limit this motor to 220hp on a hot, dry day, where a 220hp 2.5 would be much less near the edge. When considering the use of a turbo kit, there still remains the issue of getting a strong enough pressure plate on the 2.5 as a turbo engine without folding the clutch yoke in two pieces or replacing the take out bearing every 10,000 miles. This is where the WRX clutch, flywheel, and transmission case design comes into play. Without the turbo clutch setup, there is a weak link built into the program from the start. Also remember there is engine management to consider. Good "Turbo" engine management systems were one of the design features that made the 2.0 turbo motors so effective.

I still consider the use of these complete ECU "Turbo" management systems essential for a properly tuned turbo motor. It is really another complete topic for discussion, but to keep it short, the fuel and ignition curves of a turbo motor are quite different from the standard motor. The turbo control electronics ECU systems, such as that found on the SUBARU turbo motors, are also quite refined and effective duplication is frequently beyond the capability of the turbo kit manufactures, especially when trying to keep the costs affordable.

For these reasons, again, I shy away from turbo kits. That is everything I know at the moment. I remain just as cautious about just slapping a turbo on the 2.5, especially without changing the pistons down to at least 8.5, if not 8.0 compression. If one must play around with "the big power" as a friend of mine calls it, just remember, it is only money.

Thanks for taking the time to read such a long article. I hope this information has proven helpful. Thanks to Trey for kicking me into gear! I will try to ferret out more information on the 1999 engine. The big question: Is the 1999 engine the base for the rumored "2001" US turbo engine or are there yet other design changes due?

Best regards to all,

==m== Mike