Vorshlag Scion FR-S LSx Alpha Project

[Fair!]

Project Introduction - April 12th, 2014: This is the first of many posts where we will document our first GM LSx based V8 swap into an "FT86" chassis, or just an "86". That's the Scion FR-S and Subaru BRZ rear wheel drive sports coupe, for those of you unfamiliar with the nickname. Also known as "the twins", these two-door coupes are a co-creation from Toyota (ne Scion) and Subaru, and a revolutionary design for them both. These are two companies normally making plebeian FWD econoboxes (Toyota) or plebeian AWD econoboxes (Subaru) - and together they made this completely different car!

This introductory post is a big one, so you might want to go grab a snack before you start reading...



Some call it the modern day Miata coupe, which it really is in some ways, but not in others. True, it is a relatively lightweight 2 seater (there is a faux back seat) with a hard top coupe chassis, which the Miata never got, but should have. The power-to-weight ratio is similar to the MX5 Miata, meaning pretty poor. But it does give those "pure" sports car feelings, and some would argue that a "pure" sports car is always a light RWD 2 seater with lower power but good handling and gearing.

Well, we beg to differ. An enjoyable sports car is anything with RWD, good handling, good brakes, and a GOOD power-to-weight ratio. The formula is simple: Lightweight, big power, rear drive, make it turn and stop. And the FR-S/BRZ chassis almost fits that bill, but it has a few glaring holes that we aim to fill.

What's Going in and Why?

Before you wonder what we are doing in this project, I will give you the quick summary: a customer brought us a 2013 Scion FR-S with a turbocharged FA20 motor, this car had several "major engine problems" previously, and now we are swapping in an all aluminum LSx GM V8 engine and a manual transmission strong enough to take it into this chassis. This LSx V8 swap will add a lot of engine displacement and horsepower, and some much needed powertrain reliability, yet still allow for affordable upgrade path for the future. This Alpha FR-S LSx will be used for both street driving and road course use, and already has some track oriented upgrades such as StopTech brakes, 17x9" wheels, racing seats, aftermarket axles, a roll bar, and a rear seat delete.


Out with the old 2.0L engine, in with the big honkin' V8!

Before we get to the "why a V8 swap?!?" question, let's look at this chassis as it comes from the factory. This Rear Wheel Drive coupe was introduced in 2012 as a '13 model year car and has been a popular car for people that liked the small, sporty handling of a RWD 2 seater like the Mazda MX5 but wanted a roomier interior, a much more rigid chassis via the hard top coupe design, and a nominal back seat. These cars are available as the Subaru BRZ and Scion FR-S (as well as the Toyota GT86 in other markets outside the USA).

This chassis was jointly designed by Subaru and Toyota, and it shows, with some of the quirkiness from both companies survived in this final design. And a lot of "parts bin" engineering solutions, but that is going to happen on most lower volume cars like this. Some bad ideas managed to stick around, which we will address in this build. Many more details of this new-for-2013 model chassis are well documented on the interweb and we've talked about it extensively on our 2013 BRZ Project Build Thread (linked here).



In that thread we document what we felt were many of the shortcomings on the OEM FT86 models and show our numerous fixes, updates and suggestions to make them better. We show parts we bought, built, or designed to make these cars stop, turn and go better. Things like replacing the underwhelming OEM tire and wheel package, upgrading the super soft springs and suspension bushings, and fitting some proper monotube coilover dampers and camber plates up front.

Many thousands of FT86 owners have agreed that these cars work better with a few mods and the aftermarket has grown quickly for this chassis, with many offerings available to upgrade the wheels, tires, brakes, dampers, spring rates, alignment settings, bushings, exhausts, and more. An entire aftermarket segment has sprung up in the past two years to support FT86 buyers. You see it is perfectly normal to want to modify a brand new car, and real gearheads understand that virtually no car is ever built 100% perfectly for everyone from the factory, especially in the first year. The motorsports and enthusiast aftermarket exists simply because the demand to upgrade and modify cars, and the FT86 owners are no different.

An LSx Swap Is Just Another Mod

We see the LSx V8 swap as "just another modification" that can make this car better. Sure, maybe we're just taking the "mods" thing a little further than some, but here at Vorshlag we have done so many LSx V8 swaps on many makes and models of cars that nothing phases us. Not the "its a brand new car!" nor the "How can you make electrical bits work?!". We have lots of happy customers to show for all of that work over the past 12 years that we've been doing these LSx transplants. Swapping motors is fundamentally not any different than folks adding superchargers, turbochargers, or built FA20 motors.

Will swapping in a Corvette V8 motor into an FT86 piss off some purists? Sure it will. Its not "JDM, yo". But we simply don't care about any of that. We are not brand loyalists in any way/shape/form; we simply use "the right tool for the job". Brand loyalty doesn't make a car faster, and ignoring this obvious engine upgrade prohibits 86 owners from these most affordable and reliable powerplant alternative on the market. Is this "brand loyalty snob" pushback something we've run across before? Of course. The BMW community was pretty cold to the idea of an LS1 V8 in a Bimmer when we first showed it in 2002, but many warmed up after they saw, heard and rode in the "hybrid" brand V8 powered cars we created and we are quickly approaching our 100th BMW LSx V8 swap sold this year.


This V8 powered BMW E36 M3 with a Vorshlag swap kit makes 420 whp on pump gas and 5.7 Liters of LS1 goodness

The simplicity, reliability, low weight, compact size, aftermarket support, power potential, and just RAW POWER of the OHV aluminum LSx V8 is hard to ignore. Since the early 2000s these "LS1 swaps" have become very popular in virtually any car it is stuck into: the 240SX, multiple generations of RX7, various Miatas, BMWs, classic muscle cars - you name it. And we've made a good dent in the BMW purists' armor, selling so many E36 and E46 LS1 swap kits over the past 12 years. Lots of folks with street or race prepped BMWs are running around on LSx power with help from Vorshlag.



So don't look at this project and think, "What drugs are these guys on?!" There is a method to our madness, and lots of proven V8 swap packages on the roads due to the hard work we've done to develop them. There are also millions of all aluminum LSx motors just sitting around in wrecked GM cars and trucks that we can pick these motors from. This makes the LSx engine prices affordable, and with tens of millions of this engine family produced since 1996 (and still being built to this day) the market for go-fast parts for this engine is MASSIVE.

There Are Many Things That Can Be Upgraded on the FT86

We've done a lot of work on this FT86 chassis here at Vorshlag, including: developing camber plates for it, co-developing the MCS coilovers using our in-house 2013 BRZ, as well as performed track and autocross upgrades on multiple examples of 86 coupes. We've done big brake kits (however - these cars usually don't need this), pad swaps, made our own stainless brake lines, brake cooling ducts, lots of different coilover installs, bushings, swaybars, exhausts, control arm swaps, tons of wheel and tire testing, and much more. We know this chassis inside and out.



We have tracked and autocrossed multiple examples of the FT86, including former Vorshlag employee Matt's blue BRZ, shown here. These cars are quite fun to drive in autocross or HPDE/Time Trial, once you perform a few basic modifications.

continued below

[Fair!]

continued from above

We feel that the OEM 7" wide wheels and 215mm Prius tires are extremely undersized and underwhelming, but after upgrading to a 17x9" wheel and aggressive 255/40/17 street tire these things can make huge grip. The stock ride heights are also a bit too tall and the spring rates too soft for any serious track use, so we tend to go straight to coilovers (like those from MCS) and firmer spring and damping rates, then add our spherical top mount camber plates up front for alignment and handling adjustments. We've also have had good results installing Whiteline bushings and swaybars, among other solutions.



Anyway, the 86 is a neat little chassis, which can be made to handle very well with the right parts, and with an upgrade to better track brake pads/lines/fluid they stop like a beast on track as well. Read our BRZ project build thread for more on our thoughts about the brakes and suspension, which we won't really cover in this thread.


This thread is not about: suspension, wheels, tires or handling. What we are covering here is a better choice of POWER PLANT for the 86

After those non-power things are addressed, our number one beef with the Scion FR-S and Subaru BRZ has to do with the goofy drivetrain it comes with. Yes, goofy is the only way to describe the flat-four engine in these cars. The engine that comes in these things is a giant headache with very limited potential and questionable reliability when modified.

What's So Wrong With The Subaru FA20 Boxer Engine?

We aren't picking on all Subaru engines, just this one. It isn't that much of a hindrance in low speed autocross competition (these cars are doing very well in C Street and STX classes!) but the deficiencies of the FA20 quickly becomes apparent on a road course. These cars are what you'd euphemistically call "momentum cars"... since they don't make nearly enough horsepower for the weight of the chassis, once you try to get going above about 30 mph. I say the same thing about all stock-engined Miatas, among other cars. These momentum cars are better suited running with the Miata run group on a road course, as the Mustangs, Corvettes, 911s and other powerful cars are going to be flying by your FT86 (if those other guys have a good driver, of course - no amount of power can make up for a lack of talent!)



The Subaru built FA20 / Toyota 4U-GSE 2.0L "flat-4" (aka: boxer) direct injected 200 hp engine that comes in these cars is a novel piece of engineering, I suppose. It has a "low center of gravity" layout and all that, but it is also a staggering 31" wide. This is really one big ass little engine. It is so wide that the frame rails had to be pushed way out there, but even so the clearance between the cylinder heads and the frame is tight. I suppose you can remove the spark plugs without pulling the engine, but it looked tight. This layout reminds me of Porsche 911 repair nightmares (which has a similarly wide flat-6 engine).



The engine block is short front-to-back (only 18" long) but it is placed in the engine bay way too far forward. So far, in fact, that more than 1/2 of the mass of the engine is in front of the front axle centerline. This is very strange, and I cannot say why the Subaru and Toyota engineers that designed this chassis and drivetrain placed the engine so far forward. They even had to use an extra long bellhousing and a crazy long shifter extension on back of the transmission to get the gear shift lever far enough rearward to mate up with the driver in the cockpit. My calibrated engineer's eyeball says this stock motor could be moved back by another ~10", which would help the front to rear weight balance significantly; in stock form we measured 56% of the weight on the front wheels, and our Alpha tester's modded FR-S has a 55%F/46%R weight bias.


Rick's FR-S came to us with a turbo kit, StopTech BBK, rear seat removed, a 4-point roll bar and racing seats installed, weighing 2682 lbs

So the engine is super wide, and fore-aft engine placement of this lump is not great - it could have been improved with just a little bit of effort. In my opinion this car was really built from a lot of older carryover Subaru parts and technology... Subaru GR design in the back, some Subaru GD front suspension design, a traditional forward biased engine layout (things like the starter, that sits way up high and would have interfered with the firewall if it was set back, would have had to be redesigned and moved down below), and all wrapped in some pretty new bodywork. Hey, it is economics, I get it... to totally design a 100% all new RWD coupe from scratch would have cost so much more, and that would have driven up the car's price. One of the saving graces of this chassis is the low cost: The FR-S starts at $24,500 sticker and the Subaru BRZ starts at $25,495.

continued below

[Fair!]

continued from above



The engineers did do a lot of things right: it is a proper Rear Wheel Drive car, it has a rigid coupe chassis, the interior is roomy for those of tall stature, the seats are excellent, visibility is pretty good, and driver ergonomics are excellent (placement and adjustabilty of the seat, steering wheel, pedals, shifter, and other controls). Still, the specific examples of engineering compromises are numerous and the FT86 chassis contains many examples of when "the bean counters win out over the engineers". We're here to fix some of those compromises, to make a great little sports car into an EXCELLENT sports car with some serious track credentials.

The saving grace of the FA 20 engine is the height: this lump is a lot shorter than an average upright inline-4 or V-configuration engines, with the bulky DOHC cylinder heads sitting very low down in the chassis. That is all this engine does add - a lower center of gravity. But low weight isn't what this unusual engine layout offers, not hardly. As you can see below, we've already weighed the stock FA20 engine and transmission at 480 pounds. This is, by far, the heaviest 4-cylinder aluminum engine we've ever weighed.


Weight of 2.0L engine and 5-spd transmission is 493 pounds with a 13 pound dolly; the aftermarket turbo was removed

There are other things about the Subaru flat 4 engine that are unusual. With only 4 cylinders it still has an inordinate amount of parts and complexity: 4 camshafts, 2 cylinder heads, lots of cam drive chains and tensioners, and the block has to be split in half to disassemble the connecting rods and other internals. The compression ratio of the FA20 engine is also very high at 12.5:1, an uncommonly high value for a production gasoline engine. This means you have to run Premium (high octane) fuel and adding forced induction (super- or turbocharging) is a very risky proposition. The saving grace is the Toyota designed Direct Injection fuel system, which uses very high fuel line pressure and injects fuel directly into the combustion chamber (diesel engines went to this type of direct fuel injection a long time ago). These new "Gasoline Direct Injection" (aka: DI) systems tend to add 5-10% more power and 10-20% extra fuel efficiency over a more common "low pressure" (45-60 psi) port fuel injection system, and its how they can get away with that much cylinder compression. But since these DI systems are so new they are difficult to tune, upgrade and modify. Even the factory Toyota system used on the 86 includes a second set of "port fuel injectors", to help with low engine speeds and cold start problems (see this video for a better explanation of the factory DI and PI systema)

All this high tech wizardry doesn't magically make this 2 liter engine produce awesome levels of power, however. Even the 1999-era Honda S2000's F20C engine made 240 hp with 2 liters, normal pressure EFI, and 11:1 compression. The 2013 FR-S makes 200 hp with the same displacement, Direct Injection and a higher compression ratio. When measured the FA20 is making about 160-170 whp, and the chassis dyno measurement shown below on one of our tester's stock 2013 BRZ is typical.


The stock horsepower numbers are something an FT86 owner is never going to brag about

The blue test car we worked with in 2012-13 made a best dyno pull of 167 whp (measured at the rear wheels on a DynoJet chassis dyno) in stock form, and with a measured curb weight of 2775 pounds (+ 200 for a driver like me) to push around, that makes for a 17.8:1 pounds per hp ratio, which I'll round up to 18:1. Pounds per hp is a common number that racers look at, as many road racing classes have strict limits there. Our NASA TT3 prepped 2011 Mustang has to conform to an 8.8:1 pounds per hp ratio, giving us twice the horsepower for each pound we haul around compared to an FR-S/BRZ.

That 18:1 power to weight ratio is more inline with an early 1990s 1.8L Miata. For some folks, that's perfectly fine. For others, they need more power to create the higher acceleration rates that lead to more thrills, and on a road course this is amplified dramatically. Tens of thousands of people have made the comment about the 86 twins that they need more power. We've had dozens of owners ask us here at Vorshlag "When are you going to make an LS1 swap kit for the 86?" So we had been quietly looking for that perfect "Alpha customer" for the past year and a half.



Each time I have driven these cars on track (see video above), the immense grip levels were notable simply because of their inherent lack of forward thrust the 2 liter boxer engine provides. It goes fairly well, and you can terrorize stock Miatas in a BRZ or FR-S, but that's about it. We've seen so many aftermarket parts pop-up to try to rectify the limited power levels: all manner of headers, cat-back exhausts, cold air kits and other bolt-on doo-dads, custom tunes and even E85 ethanol fuel conversions. All sorts of manufacturers and shops have been trying to unlock more power - but most solutions seem to come up short. There just isn't a lot of "easy" horsepower left to untap on the FA20 engine, and many of the claims of "300 whp!" end up being a bunch of internet hype or turbo ticking time bombs. Even a 200 whp FT86 build is somewhat legendary.


Weight of a bone stock 2013 BRZ Premium package car came in at 2775 pounds on our scale

Why I Hate Boosted Track Cars

The main power solution now coming to market for this chassis includes forced induction kits using either a turbo- and supercharger. I lump both of these Forced Induction (aka: "FI") solutions into the same category, which I often label with the derogatory term "boost buggies". This Alpha build FR-S came in to our shop with an turbo kit installed, and was one of the first Ft86 car's in Texas with this particular kit. I don't know what brand of kit this is, but it almost doesn't matter - as these FI kits will all suffer the same fates in the 12.5:1 compression ratio FA20 motor. Detonation and premature engine failure until the engine is modified with new pistons that can lower the compression ratio.


When the lowest part of your track-built car is the turbocharger and engine oil line, trouble awaits

The layout of this particular turbo kit was less than ideal, in my opinion, especially for track use. The turbocharger was mounted low, underneath the engine, and both the turbo housing and the oil feed line to it were the lowest parts on the chassis. There was also an electric oil pump for the turbo oil feed. This would mean than an off-track excursion could easily rip the oil line off or damage the turbocharger first, which could turn into an expensive mistake. This car was supposed to be built as a reliable track car, but this kit pretty much precluded that, as you have had to make the bottom of the car safe for "offs" by adding skid plates, thus making ground clearance even worse. This turbo package and associated tuning also ended up killing two motors - first the stock 12.5:1 motor then a lower compression motor - both of which died on the dyno, according to what we heard. Before the final race prepped turbo motor was ever able to take a single lap on track, the car's owner went looking for a better solution.

continued below

[Fair!]

continued from above





There's not going to be a blame game here - it was a commercially available turbo kit, but something obviously went awry. Several times. Even without multiple engine failures on the dyno, I think adding a turbo system to a high compression stock motor with cast pistons was always going to be a recipe for disaster. Maybe it would have lived with the final race built motor it received, but we will never know because that motor is no longer in this chassis. There were too many other things wrong with the installation to document, but needless to say, we're going to put it back together the Vorshlag way - with a big naturally aspirated V8 making a reliable 400 whp, a beautiful exhaust system that has a sound that will make you crave it, proper cooling and sub-system solutions, and a distinct lack of zip ties and bailing wire.



A Pikes Peak team we sponsored with some suspension bits had decent luck with a turbocharged set-up on their FR-S race car, but it had some serious tuning hiccups and teething pains to deal with before it got to the mountain. For that extreme hill climb you see over 14,000 feet of elevation at the top of the 12 mile race course, and the only way to maintain horsepower at the top of the mountain is with FI or electric power. FI can make up for the lack of air density and electric motors don't ingest air, so they don't care.

But most of us don't race at 14,000 feet, and in most cases a larger displacement naturally aspirated engine is much MUCH more reliable than any FI engine, especially engines that were not FI from the factory. I'm personally not a fan of turbo and supercharger "kits" installed onto stock motors, and I have seen and experienced a lot of bad examples of this first hand. I worked at a tuner shop that specialized in supercharged V8s back in the day and customers were always breaking pistons, head gaskets, and more. The only way to make them live, even on drag/street cars, was with very mild boost levels + a GOOD fuel system + a conservative tune + a very well built motor. That hasn't changed.


Vorshlag has worked with FI engines for road course and hill climb use and we understand the REAL challenges to making these reliable

The problem is exacerbated when the car/motor in question never came from the factory as a turbo model. You see, there is a LOT of engineering and development that goes into building a factory turbocharged car, with all sub-systems getting upgrades to deal with the extra power and HEAT of a boosted engine. To make a non-turbo car into a turbocharged set-up is a lot of additional work to get it reliable. It is so much more than "bolting on the turbo". All sorts of things need to be insulated and re-routed, all fluid systems need beefed up cooling, and heat soak becomes a major issue.

And don't get me started on what it takes to make a turbocharged or supercharged engine car live on a road course. The things you need to make a turbo engine live in drag racing or street use pale in comparison to what you need to do for track use. Everything heat soaks... the turbo housing, the intercooler, the intake manifold, the oil gets cooked... you almost cannot make an intercooler big enough to shed the heat from a turbo to run on a road course for more than a handful of minutes. The intake air temperatures on most turbocharged engines used on a road course keep going up, up, up.... to the point where the pistons eventually melt or something goes BOOM. Even short road course use set-ups like Time Attack and Time Trial face huge hurdles and challenges to keep the motors together and reliable.


Road course competition in any FI car includes a lot of extra prep and maintenance

So we've explained why an FI build doesn't make sense for this dual purpose street/track car. What is our solution? If you've heard of Vorshlag then you already knew what we had in store before you read anything on this post - swapping in a lightweight V8 engine that will nearly triple the stock displacement. That is a common answer we have to the question of "I need reliable horsepower". Our plan is to use this particular 2013 FR-S as our "Alpha" or first prototype build, then perfect a bolt-in LSx V8 swap kit for FT86 owners to be able to swap in this aluminum V8 motor with as much ease as possible. We always start all new swaps with an Alpha build, which includes the following cars...

• Vorshlag Announcement Thread for all LS1 Swap Development
• BMW E36 LS1 - one of many build threads
• BMW E46 LS1 Alpha
• Mazda NB Miata LS1 Alpha
• First BMW Z3 LS1 Swap
• BMW E30 LSx swap

Vorshlag is well known for LS1 swaps in chassis, as well as our production kit components that make installing these V8s easy

If you do some research you will see that Vorshlag has developed and built LS1 V8 swaps for a number of different chassis, and our E36 swap kits are extremely popular with new sub-system parts and products released from 2007-2014. We released our initial "Stage 0" kit (the parts necessary to mount the drivetrain) for the BMW E46 chassis last year and those have also been received very well. We have done a number of other "one off" LS1 swaps such as the BMW E30 LSx that we built for the Grassroots Motorsports $20XX Challenge, which we won overall with in 2011. Then there was the BMW Z3, and we're also need into the development for an NB chassis MX-5 Miata V8 swap kit.


The beautiful engine bay of the Z3 LS1 was worth opening the hood to see!

continued below

[Fair!]

continued below

continued from above

Alpha FR-S LSx Swap Outline

What we have planned is simple and meshes with the customer's needs and the shop that brought him to us. I will give the basics here, and show the step-by-step fabrication and development pictures as the V8 swap kit is developed. We won't give away EVERY detail, as we've had copycats try to duplicate our work over the years. So I won't give away part numbers, finite measurements and the like, but we will show more than enough to let you readers see how we do a new Alpha build.



The owner of this car is named Rick and he found his was to Texas Track Works in late 2013, a shop in Ft Worth the specializes in track prep, race alignments and more - and is a Vorshlag camber plate dealer. Even though in some ways we are competitors we are we are friends with the staff at Texas Track Works and will partner with them to complete this Alpha build. They are the reason why we have the opportunity to work on this particular FR-S, and for that we are thankful. Rick rode in one of TTW's customer's track/street cars: a Nissan 240SX (above left) with a LS1 swap installed by TTW. We've raced against this car in NASA TT3 in our TT3 Mustang (above right) and that Nissan really scoots. This Nissan is a clean install - unlike some of the hack V8 swaps we've seen built elsewhere, the TTW crew gets the details right and has the race experience to know what matters. After that test ride in that lightweight RWD coupe with proper LSx V8 power, Rick knew what he wanted in his FR-S... LSx V8 power! After that decision it didn't take Texas Track Works long to make the call to us here at Vorshlag, and we came to an agreement: where Vorshlag would cover the development costs of this swap if we could see a way to make it as a kit. So they brought us the car in early January 2014 and we took a look...



Once the OEM drivetrain and turbo kit were removed we got a closer look at the naked the engine bay. Wow, the underhood area is HUGE! The frame rails are a staggering 31" to 33" wide (they taper in at the back), to house the immense Subaru flat-4 cylinder engine, and I felt immediately that the LSx swap would fit well. We first took a LS1/T56 mock-up drivetrain we keep on hand for swap development and tried to stuff it into the car...


Our first LSx V8 + Tremec T56 mock-up had the shifter portion was way too far forward, but we knew what to do.

Hmm, the traditional 1998-2002 Camaro/Firebird Tremec T56 6-spd was too short, as was the next drivetrain we tried: an LS1 with a Tremec TKO600 5-spd. That transmission length was even worse, and no matter how far back we moved the engine the shifter location was still way forward of the factory shift boot and opening in the center console. The TKO600's shifter was way inside where the dash would be. Neither would work.


These two secondary braces had to be removed, and after using our spot weld cutter on 6 spot welds each one fell right out

After removing one spot-welded transmission tunnel brace (see above) and unbolting a few items mounted to the firewall, we did some more drivetrain mock-ups and got the engine back as far as we could without serious tunnel surgery, something we always try to avoid in any swap we want a customer to be able to replicate in their home garage. We measured the shifter location on a ZF S6-40 6-spd transmission we also had on hand, and it wasn't anywhere close either. We finally called our transmission expert, Joe D, and he mentioned the Tremec Magnum XL 6-spd that featured a direct mounted shifter about 5.5" further back than the normal direct-shift T56.


The second drivetrain we swapped in was another LSx V8 and a Tremec TKO600 5-spd

continued below

[Fair!]

continued from above

This transmission was developed for the S197 5.0 Mustang chassis as well as the 5th gen Camaro LS3, but never came from the factory installed in either car. Both of these cars come with factory transmissions that have remote mounted shifters, and in both cases the shift feel is terrible. We've installed T56 Magnum XL 6-speeds into late model Mustangs and the difference in shift feel is remarkable, not to mention the extra torque capacity these have (rated at 700 ft-lbs of torque!) compared to the stock Gretrag MT-82 in the 5.0 Mustang.


The long-awaited and custom ordered Tremec T56 Magnum XL arrived!

The problem was, Tremec hadn't yet made the production model of the transmission to work with the GM LS-series engines. Joe said he could custom build us one with an input shaft made to fit the flywheel depth and pilot hole diameter of the LSx, like he has done for Grand Am and World Challenge racers with 5th gen Camaros. So after many measurements and inspections on the FR-S engine bay and tunnel we took a gamble, ordered this custom Magnum XL transmission and waited for it to arrive. We also picked up an LS2 GTO 6.0L oil pan, which has a "front sump", hoping it would clear the FR-S crossmember. Based on previous swaps and where this FR-S crossmember was located, I felt this "front sump" oil pan would fit.


The final LSx + T56 Magnum XL drivetrain placement was about as perfect as you can get

After many weeks of waiting the Magnum XL arrived and we got to work on the final series of drivetrain mock-ups. We also bought a new LS3 intake manifold kit, complete with fuel rail and drive-by-wire LS3 throttle body. We want to use the later model Corvette LS3 PCM (power control module), as it has a faster processor and advanced CAN-BUS options, to better match the FR-S systems. That means we would use the LS3 intake manifold, which has an electronically controlled throttle and not a direct cable controlled throttle. The cylinder heads for the LS3 are different than what we have on our mock-up block, but at this step it didn't matter.

After about 6 or 7 iterations and tweaks with the "LS3" + Magnum XL drivetrain in and out of the car we finally have the engine where we want it, as far back as the heater core and intake allow. We then tweaked the driveline angle (downward tilt of the engine) to match the factory diff angle. Then we started building the prototype set of motor mounts.



That's where we are now, and today our crew removed the interior to help develop the transmission crossmember. After those bits are done we will develop a full length 1-3/4" primary stainless steel header with 3" collectors. Unlike what some other V8 swap kit makers like to do, we make real exhaust headers the hard way, because it makes SO much more power everywhere. The shortcut normally taken by others it so buy a cheap set of 1-1/2" "block hugger" shorty universal fit headers, heat them with a torch and smash any tube with a hammer that gets in the way.

continued below

[Fair!]

continued from above


The BMW E46 LSx prototype headers were developed and built in house, with final production by a major US header manufacturer

That short cut header is a huge restriction and can cost upwards of 50 hp, but making custom full length headers is so much work that most shops won't bother. We wouldn't have it any other way, and our production full length headers tend to add 40 hp over stock. Once the prototype headers are made here at Vorshlag we will ship them to our production header supplier, where they will make fixtures based on our prototype set. Then they will make a production run of stainless steel headers using a CNC bender, making our kit headers both more cost effective and improving fit repeatability.

And yes, for you California folks we have tested many sets of OEM exhaust manifolds and found one that fits, so this swap could feasibly meet 50 state emissions and pass the CARB referees, but we won't likely do one of those for a while. Our installs will all have our headers, and if its a street car we will add proper catalysts as well.


Vorshlag production 1-3/4" x 3" full length stainless steel headers fit the E36 BMW chassis like a glove

Everything else will be built in-house, like our production E36 and E46 motor mounts are now. We use CNC laser cut flanges, bolt them into our production fixtures, then TIG-weld everything together. Each welded part is testing on another fixture, then they are powder coated for long term durability and good looks.



After the drivetrain mounts and headers are built there are still a dozen of other parts and sub-systems to tackle, but its nothing we haven't done many times before: Cooling system hoses, reservoirs, radiator and fans; intake air tubing, filters, and MAF sensor; electrical wiring to the engine, gauges, OBD-II port and CAN-BUS integration; driveshaft, shifter and transmission crossmember fabrication; fuel pump, lines, hoses and regulator, from the tank to the engine; and last would be the relocation of various components that were in the way of the V8. There could be more than that, but so far that is a valid list.

What About the Weight???

The number one question we get on every build is "how much weight will the V8 add?" Don't worry... we know Newton's Second Law of Motion: F = ma. We're going to really up the "a" with the LS engine and keep the "m" as low as possible.



We here at Vorshlag are fanatics about weight, even when we are racing in our 3800 pound TT3 Mustang, heh. We weigh everything during a build like this, and we have already weighed lots of FT86 bits in the BRZ project build thread. You will see lots of pictures with various bits on one of our digital scales in this thread also. Why? Because the internet is full of poor data, numbers that turn out to be bogus, and people that are full of crap. We bring the tech, and when it comes to weights we show pictures of anything we reference on a proper scale.


Here's an example of a race-prepped BMW 3 series with LS1 V8 power at 2508 pounds. That's with a full road race legal roll cage, too!

Often times, as with our BMW LS1 swaps, the additional weight of the all-aluminum and compact LSx series of V8s is nil, and we only gain weight in the transmission (the T56 6-speed is a brute and weighs over 125 pounds). Our full street duty/full interior BMW swaps are usually within 20-40 pounds of the stock set-up. The iron block inline-6 engines used in E36 3 series BMWs is HEAVY and long, and the LS1 engine is a hair lighter and about 12 inches shorter, which moves the drivetrain mass rearward for better front:rear weight bias. We feel that the FR-S LSx is going to be much along the same lines, with more rearward weight bias and a gain of 40-50 pounds, at most. This isn't some wild guess, but based on dozens of previous swaps, and weights we've taken of these engines before.

Remember the 480 pound measurement of the factory FA20 and 5-spd, shown above? Well a Tremec T56 Magnum XL is 128 pounds and a factory LS6 weighs 457 pounds, with a lighter than stock flywheel (which we always do - it doesn't need a 48 pound 2-piece factory flywheel, even for street use) plus the heavy OEM exhaust manifolds and a power steering pump. This 86 chassis has electric assist power steering built in the steering column, so we will remove the hydraulic power steering pump from the LSx engine we end up with, and our exhaust headers always save weight over OEM cast iron manifolds. After we do the math it looks like it is going to be pretty close to the stock weight, which is good. The weight of a turbocharger and intercooler from our before weight here is not insignificant, either. If we're wrong on that guess, we'll show it here, too. Look for the final weight of the FR-S once this V8 swap is done, which we will compare with the "before" weight.

What's Next?

That's enough for now. Tune in next time when we have the motor mounts completed and start to tackle the next bits like the driveshaft, headers, fuel system changes, and wiring. And yes, we have an OBD2 expert already on the case to make the GM computer talk to the factory FT86 computer, and our build here will have functional gauges and electric power steering.

Please note: ANY NEWS about the swap kit, swap parts availability and pricing will be posted here AS SOON AS WE HAVE IT. Please don't call our over-taxed employees at Vorshlag or the folks at Texas Track Works to ask "when will the kits be released??" - we will post any news here first. :)

Thanks,

[eliandi]

Really enjoyed this series, Terry. Belated thank you for having SCCA out to your shop for tech. It was like Santa's workshop :)

[Fair!]

Really enjoyed this series, Terry. Belated thank you for having SCCA out to your shop for tech. It was like Santa's workshop :)

hehe, Santa's Workshop, I like that. We make the toys for gearhead girls and boys...

http://www.dfwmiata.com/threads/2591...-Alpha-Project



We love doing this SCCA Tech/Open House every year. Once we move into our larger facility (later this year) we'll try to do this more often.

[SixAce]

cool story bro

[hornetball]

May God-bless normally aspirated, big-displacement pushrod V8s! Compact little packages of reliable torque. Is it any wonder MB and Nissan are struggling in the V8 Supercar series? They're missing the pushrods, LOL.

Every time I pull a Miata engine, I marvel at their physical size. Big heads make for big, heavy engines. Don't even get me started on the POS 3.8L MB V8 in my wife's 380SL.

I'm looking forward to seeing the taillights and hearing the exhaust note from this car as you rapidly pass my 1.8L Miata. I'll fix that someday, but in the meantime I'm working on the nut behind the wheel.

[Sloaner00]

Great writeup. I enjoyed your attention to detail and look forward to more updates.

[Fair!]

Project Updates for June 9, 2014: It has been a hectic couple of months at Vorshlag since my last update. We were buried in early 2014 season preparations and parts sales for many of our customers as well as prepping and racing one of our cars (2011 Mustang in NASA TT3). We also have to pace our V8 swap kit development projects like this one (where Vorshlag writes off about 75% of the hours towards kit development) with paying customer service and race prep work - that pays the bills. I got a little busy myself as well, but after a nasty 150 mph crash I had at Road Atlanta a month ago, that side-lined me from racing for a bit, I'm getting caught up on a lot of project build posts like this one (and this, and that, and even this). During these past two months our shop has made some good headway on this FR-S V8 swap and the NB Miata LS1 Alpha project, which I'm writing an update for next. Let's go ahead and dive into where we are now on the FT86 V8 swap...

Drivetrain Mounting in Kit Form is Tricky

As I've said before, making a one-off V8 swap is relatively easy. Bring us a car, choose the drivetrain, and we can get it in there in a couple of months for a race car (we did just that for a customer in January-February, as well as full race prep, custom brakes, custom cage, all safety gear, and a full re-wire and re-plumb from front to back). One-off V8 swaps go much faster because you can cut things out of the way and just fabricate the best solution in the tie allotted. Making the custom motor and transmission mounts even for a one-off V8 swap is still relatively easy - its takes about a day and a half on something like this FR-S, once you have the drivetrain in the right spot.


Left: Big ugly hole in firewall patched with aluminum, held in place temporarily by Clecos. Right: A big hole was filled with shiny new steel

We could just stuff the drivetrain in there wherever we wanted; cut away the firewall or tunnel or whatever is in the way; measure everything to get the driveline angles perfected, leveled and centered; then cut a hole for the shifter. After that was done we would go back and plate over the holes with aluminum or even patch in sheet steel to make it look OEM, like we did on the firewall for this one-off Ford Coyote 5.0L V8 swap above last week. This was a one-off V8 swap that was brought to us to complete. To fill some holes in the firewall we used two methods: aluminum panels in the left picture and seamlessly integrated welded steel panels on the right, as shown above.

Making these custom panels to cover up holes or other sins is common on a race car, and they can be done on a street car and look seamless and factory, but not easily in kit form. The pictures below were on a car that never had this particular transmission installed by the factory, and it didn't fit the trans tunnel. So part of the tunnel was cut away then a new cover was custom fabricated out of aluminum and bolted in place.



With the finished transmission cover panel bolted in place (with captured "rivnuts" so you don't have to hold a nut on the back side) and later covered in thermal insulating material, it was no longer an unsafe gaping hole in the floorpan. This particular aluminum panel structure can be unbolted to access the tailshaft, add fluid to the shifter opening or replace/re-grease shifter mechanism. It also makes drivetrain removal even easier when removed, yet it is still heat and fireproof. On a street car this could all be under carpet and fully functional.

But for an engine swap that we want others to be able to replicate in their garage using our parts, this type of significant tunnel or firewall surgery and custom fitted panels is not an option. You don't want to have to try to copy where we cut out this piece of sheet metal then make patches to fit, no no no. Yet there are some popular LS1 swaps out there that do expect you to strip your street car down to the bare tub, cut out the transmission tunnel or firewall and just "make it clear the engine where our mounts put it". Significant welding and custom fabrication work is not what we'd call a "do it yourself swap kit"



Our BMW swap kits are made to bolt-in without any welding or fabrication needed, and that's how we're designing or FT86 LSx swap kit - as much as possible. To make a swap kit that is easily repeatable takes a LOT more work involved in the initial design. Our choices of drivetrain parts and locations are narrowed considerably when you cannot cut the tunnel or firewall out and replace pieces or entire sections with new bits. You have seen how many transmissions we've tested on this swap (4). We could have made any of them fit with enough custom fabrication work on the tunnel and firewall, but that's not a "production swap kit", that's a "one-off swap". And we are making a production kit here...



The remote coolant reservoir above, which we make for our BMW E36 chassis LS1 swap, bolts into 3 factory threaded holes in the chassis. It has an integrated mount for the factory heater control valve, relocated from across the engine bay. It houses the factory BMW level sensor on the bottom side, which fits in a channel that happened to be there. The heater hoses clear the routing for our A/C lines. Making new parts fit with all of these constraints takes lots of extra time to plan for and build around. We have to make all of our swap kit components key off of and/or bolt into factory holes. This ensures everyone can locate the drivetrain or other bits we are changing in the same exact spots as we designed. We also go to great lengths to prevent the need for a welder, or cutting off of huge chunks of the chassis, firewall and transmission tunnel.

This extra work required to make everything function as factory looking swap, with proper engineering principles, takes tons of extra time. Time we get to pay for, not the Alpha customers that work with us. They just have to be patient while we go to these extra lengths, on our dime, but our Alpha swap customers have all been great.



The way we make the mounts also has to take into account production manufacturing techniques, like computer CNC Laser Cut portions of sheet or plate steel or aluminum. We even add things like holes, slots and tabs to key or bolt together during fabrication, using our custom production fixtures, to have make each motor mount or transmission crossmember to be as accurate to the original. And fit the chassis as well as as possible. Making a one-off set of mounts would be SO much easier.

continued below

[Fair!]

continued from above

FR-S LSx Drivetrain Mounts Version 2.0


The more we looked at the hood clearance, the less we liked the old LS3 intake and throttle body location. It was time for Ver 2.0

We have already made one set of motor mounts for this swap, but after closer examination, I had the guys chunk them in the recycle bin and start over. We had left an extra 1/2 inch that we could move the engine rearward after removing one additional, bolted-on factory part off the firewall. That meant 20+ hours of measuring, mock-up, re-measuring, tweaking and fabrication were thrown away. Moving the engine back as far as possible without cutting the firewall is THAT important, especially when we could make a lot of these FR-S swap kits (we've officially surpassed 100 kits for the LS1 BMW E36 chassis now).

Our crew here was cool with it, just more development hours. Once the engine was moved back to the final, optimum location Olof and Brad noticed one or two factory bits that were really in the way of a better mounting solution. Our new head fabricator Ryan H expressed his concerns over a couple of stamped sheet metal bits that he thought should be removed, so we had a meeting and after a few spot welds were cut, the remaining tunnel structure was much more favorable to the mounts we wanted to build.



As you can see in two pictures I originally posted in my first update, shown again above, the Tremec Magnum XL was getting awfully close to one of the transmission tunnel support brackets. This is a U-shaped stamped sheet metal structure spot welded into the sheet metal tunnel that the OEM transmission mounting bracket supports were then spot welded to (see my original post describing the removal of those). After we moved the drivetrain back that crucial 1/2 inch this U-shaped piece interfered with the Magnum XL. Instead of having our customers just "clearance here" on this structure we decided to simply remove them. And in any case, after we had removed the two little brackets (post #1) this big U-shaped support was no longer needed (it was simply a place to mount other brackets to).


With the U-shaped structure finally removed (about two dozen spot welds) the transmission tunnel had lots of room

continued from above


Our 2nd set of motor mounts and new drivetrain location required removing these 3 spot welded brackets

Above you can see the U-shaped brace that was removed, as it no longer served any purpose and was just in the way. At right are two little funky, angled, bent steel pieces were barely attached to the steel front subframe welded assembly and after a couple of spot welds they were gone.


Left: Our first set of mounts had weird bends. Right: After removing two stamped steel bits our motor mounts will fit a LOT better

Removing these slotted black brackets removed a big angled protrusion from both spots on the subframe where we wanted to land our mounts. The first mounts we built had weird bends to try to go around the factory bits (above left), but now with a nice square section of crossmember to mount to (above right) the new lower portion of the motor mounts will sit flush against this vertical wall of steel - and still bolt into factory drilled holes. This makes our new LSx motor mounts (version 2.0) a little easier to make, but it will now have an additional hole to secure them to the subframe and sit more securely there. Removing these 2 brackets was super easy - 2 spot welds per side and they popped right off.

continued below

[Fair!]

continued from above



Here is the U-shaped transmission tunnel structure that was interfering with the big Tremec Magnum XL 6-speed manual, so after some spot welds were cut it was removed. There were a decent number of spot welds, but it came off in about 30 minutes. See my quick write-up for Spot Weld Removal tips below - its fairly easy and takes no special skills, just a $5 spot weld cutter and a drill.



Olof made the new Tremec Maxnum XL specific crossmember fabrication (above) that squeezed into the tunnel nicely. Two side plates will bolt to and key off of the chassis, then have 4 holes per side that are drilled into the transmission tunnel (see them bolted in place, two rows above). Then our new U-shaped transmission crossmember will bolt to those side plates, which have some bolt-together attachment points integrated into them. One of our Vorshlag bolt-through 95A polyurethane T56 mounting bushings will go between the transmission crossmember and the Tremec 6-speed.



These brackets and mounts have since been finished welded, templates were made and reverse engineered from each piece, and CAD drawings produced for laser cut production runs of these mounts and crossmembers. I'm happy where the drivetrain ended up, as this gained us more room at the throttle body/hood and moved the weight back even more towards the rear.



Here are the semi-finished prototype motor mounts and transmission crossmember pieces that we will use for the LSx V8 and Tremec Magnum XL. Unlike some shops that never show off their swap parts, we're showing the unfinished prototypes. I'm sure some of you arm chair engineers will want to comment, but remember: 1) these are prototype parts and 2) we have made 100+ V8 swap kits over the past 8 years and have had zero failures. We still have a few more gussets to add to the transmission crossmember mounting pad, then these welded parts will be painted, assembled with the correct poly bushings and hardware, and installed back in the car next week. They are light and strong, and we will road/track test them before going into production for anything.

continued below

[Fair!]

continued from above

Cutting Spot Welds Is Easy

This V8 swap kit installation is going to require a little bit more work than some of our swaps we developed in the past, but it is nothing insurmountable by any able bodied person with a drill. There are 5 factory stamped sheet metal pieces we have removed - three from the tunnel and two from the factory engine crossmember. These are attached with simple spot welds. It is unfortunate that we need these removed for the kit to work well, but it is worth it for a better engine placement.

In this section I will show the process for removing two pieces of sheet metal parts that are spot welded together. I did an entire roof panel removal in my home garage, which had hundreds of spot welds, and it took less than half a day. These 5 pieces on the FR-S will take even less work.


This "roof swap" wasn't much fun, but all of the hard work was in the welding. Removal of the old skin was easy

Several years ago I owned a BMW 330 Coupe that had a sunroof and full interior, which I raced in an SCCA class (DSP) that didn't allow for gutting the interior or adding a sunroof delete panel (there are carbon fiber panels you can bolt or bond in place of the sunroof cassette). I couldn't sit in the car upright without my head in the roof panel, so I converted the BMW to the factory non-sunroof structure for more headroom. This conversion included a new roof skin, an addition of one more roof bow, and a new headliner. We started by removing the entire roof skin that was both bonded and spot welded to the surrounding structure in about 200 places around the perimeter. Some of these spot welds were hard to get to, as we had to remove the front and rear glass and all of the interior to even see them. But once we had access, we had the roof skin off in a couple of hours using a drill, a spot weld cutter and some elbow grease.


Left: Typical spot weld cutting tool (small hole saw with a spring loaded centering pin). Right: Spring loaded center punch marks the spot

A spot weld cutter is essential a tiny hole saw with a spring-loaded alignment pin. The goal is make a small dimple in the middle of the spot weld, align the pin in this dimple, then use the tool to cut around the spot weld. This releases the top layer of steel from the base layer. If you do it correctly only the outer layer has a (1/4" diameter) hole in it - and this is the layer we want to discard - while the base layer ends up with the little "bump" of extra weld material left. This bump is the weld and steel from the outer metal layer, which can be ground smooth or in this case, ignored.

On the FR-S, we're just removing some things that are no longer needed to make room for the new transmission or engine mounts. Luckily we don't have to weld anything in place of these removed parts, or even grind the remaining bumps smooth (that's optional - and we did that for better aesthetics). These stamped steel parts are spot welded together without any bonding agent and come off with relative ease. Sure, the joints have a bit of seam sealer smeared over the edges, to prevent water from getting underneath them, but with a little coercion they still come right off once ALL of the spot welds are gone. The spot welds are easy to see - each one leaves a small round blemish in the outer layer of steel.


The spot weld cutter is simply a tiny hole saw with a guide pin that aligns with the punched dimple you add in the center of the spot weld

The FR-spot welds we were after are all easy to see and easy to access, with the drivetrain out. Only the outer layer of sheet steel needs to be cut with the spot weld cutter - the inner transmission tunnel metal should be left alone. Once the spot welds are gone from the outer section these stamped parts can be pulled off with your hands, but if you get a stubborn spot weld (or don't drill it from the center) a chisel or screw driver can be used to pry the metal apart.


On this car there was a bonding agent in addition to the spot welds, so they had to be pried apart. On the FR-S the panels were only spot welded

Rest assured, this step is not too difficult and you won't be welding things back onto the car once these bits are removed. Just be prepared to spend a few hours with a $5 spot weld cutter and a drill before our production motor mounts and transmission crossmember will have enough room to bolt into place. We will show more of this step after the kit is in production, with some videos.

What's Next?

Right now the FR-S is sitting with the drivetrain temporarily held in place (below) by an engine hanger while the prototype motor mounts and trans crossmember are being welded, painted and assembled. Then we can get the final engine in place and gather the parts needed for the engine accessories (Corvette LS2 or something compact like these). After the engine is fully assembled and the radiator is back in place we can finally begin prototype header construction, a custom 1-piece aluminum driveshaft can be measured and built, and a cold air inlet tube can be design and routed.



In our next installment I'll show these steps listed above and detail the final motor being used for this car. It will likely be an LS1 5.7L with an LS2 intake manifold in this Alpha car, but since many folks doing this swap will opt for the "easy button" 430 hp 6.2L LS3 crate motor from GM (about $6500, with a warranty) and the includes LS3 intake manifold, we will make sure both LS2 and LS3 variants fit. While there are no external differences in any of the LSx V8 family of blocks/heads, these two intake manifolds (LS2 vs LS3) have small changes that we need to design around. The LS1/LS6 intake manifold are very different (and have a mechanical throttle body, which we don't want; this car has an electronic throttle pedal) and push the throttle body way too far forward, but we'll test and show that in the next build thread update.

More soon!