Last one standing

So having previously discussed all of the other radiator placement solutions for a T2b that was originally air-cooled, there's only one place left to go: below.

This is a first generation solution (circa 2005) from
Custom Off Road in Queensland, Australia. (Now Custom Vee Dub)
The radiator body is fed air by forward motion via the
louvers punched in the bottom plate. Electric puller fans above the
assembly come on when needed.

If you've ever examined this idea, you'll find there are a lot of mixed emotions out there about it.

• "A Radiator does not belong UNDER a car."
• "How is this any more efficient than the other ideas proposed?"
• "Aren't you concerned about bashing up your radiator and spilling all your coolant?"

But I've been chasing this bear around by the tail for the last few years. Lots of folks have put their radiator underneath and the success rate is no more mixed than radiators in the engine compartment.

Unlike the Vanagon conversion, however these underbelly conversions didn't get their start in the sand and surf California scene. SoCal is ground zero for all of the continuing air-cooled scene. If you want to stay stock, most of the innovators (ha!) in Type1 (Beetle, Ghia, Split-window Microbus) are there. But the folks in SoCal are crippled by their own emissions regulations. And the later the vehicle gets, the loopier the laws are. So there is precious little innovation coming out of SoCal these days for the late bus owner.

The source of the most of the really aggressively engineered ideas are coming from Australia and the UK. Engine conversions are not at all unheard of down under, and the appetite for the 'Caravan' in the UK is undimmed, to the point where the Brits are buying dry Buses from SoCal and shipping them to the damp island.

With all of this motivation, is it any wonder that they were the first movers?

The mesh at the front to keep from scooping up
road kill is a nice touch. But with your fans on
the bottom, you're sort of asking to get them
ripped off the radiator. Better protection is
necessary though at least these
are mounted as puller fans.

Unfortunately placing the radiator underneath has raised all kinds of differing opinions about what the best way is to provide airflow to such a beast. There are heated (no pun intended) debates about 'Bernoulli effect' and push fans vs. pull fans, and whether to put the ECU in control of the fans or whether to run the fans as their own subsystem, with the temperature sender acting directly on the fans and telling lies to the ECU.

But the one constant is that, to be sure to run cool, you'd better have a scoop, and you'd better have fans. This doesn't mean that the scoop has to scrape the ground. The T2b has a startlingly high stock ride height, officially 200mm/7.8inches. Yet that actually describes the bottom of the front suspension beam e.g., the lowest part of the vehicle. The T2b (and the T1 and T2a before it) are built in a 'ladder frame' construction. The bottom of the actual frame rails (the bottom of the frame of the vehicle) is 300mm / 11.81" almost a foot off of the ground! And that doesn't include the extra 100mm (about 4 inches) from the bottom of the frame rail to the actual floor of the platform. (There are some necessary occlusions, like the shifter rod tube and the clutch cable tube.) Indeed, there is actually gobs of space to use...if you use the right radiator.

The rule of thumb for sizing the radiator is that it needs at least the same BTU capacity as the vehicle that the engine originally came from. In some cases, people have used two radiators plumbed either in series or in parallel, and still experienced overheating. Why?

The use of parallel radiators has shown itself to be a problem of indeterminacy: for what ever reason, one radiator has less resistance to flow than the other. So all the flow goes there until the resistance equalizes. But all this means is that one radiator does most of the work, and the other one stays mostly cold, even if the other side of the coolant loop is so hot that it is cooking the engine in its own juices.

Series radiators do better, because all of the coolant is forced through both radiators. But compromises have to be made: if you plumb the output of radiator #1 to the input of radiator #2, you have to make sure the connection curves up higher than either radiator and that there is a radiator vent valve there to vent the air out. You can wind up with a blown head gasket for failure to get the baby 'burped' correctly.

So ideally, you'd want a single radiator of the total cooling capacity you require. Additional plumbing has the potential to be headache enough. Why make it harder? Stick to a single radiator that will fit the space, and meet or exceed the BTU capacity of the stock radiator.

The issue of a scoop is enough of its own issue to merit a separate installment on this blog, but one thing has become abundantly clear to me: tilting the radiator itself is not a wise design move, and tempts you to other follies like not ducting the radiator properly or to placing your fans on the bottom of your radiator where they can be sheared off. The following design practices are required to make a good radiator installation:

• Air that reaches the the radiator must do so under positive pressure. There are lots of different ways to achieve this, either by setting the radiator upright at the front of the vehicle (no matter how ugly it looks) or by ducting air to the radiator, but no matter what, it needs to be under positive pressure MOST of the time, e.g.when the vehicle is moving.
• Air that is leaving the radiator needs somewhere to EXIT. If you duct from the bottom up, once the heated air is in the cavity, it has to have somewhere to go. In an ideal world, that cavity would vent to an area of lower pressure, thus causing the heated air to move from high pressure (the duct) to low pressure (the cavity) before being exhausted out the area of flow. A Venturi effect.
• When you use a fan to make up for a lack of sufficient input air volume, PULL, don't PUSH the air through the radiator.
• Box in the assembly. You can't make any of these air pressure or air speed effects work in your favor (and you need all the help you can get, considering the compromised placement of the radiator) if you just screw the assembly to the bottom of your bus and call it done. You need these passages sealed. Remember rear radiator solutions? The air will find the passage of least resistance, right past your radiator. The same thing goes for your ducted installation: seal it up!

A special note: I've gone all 'Father knows best,' here.That's not because I do know what's best, I've only just observed others closely and seen what has worked and what hasn't. You may be able to use a massive enough radiator and enough fans to cause your T2b to lift-off from the ground! You might even have enough oomph with this unducted design to keep it from overheating. But the four principles above all but guarantee success. If your goal is to drive and ignore the engine temperature, designs based on these attributes can't help but work.

A million zany ideas

Since the advent of water cooled engine conversions for the VW, there has been the perennial problem of where to put the radiator. There are three generally acknowledged places to put it:

The Brazilian T2c imported to the UK with the radiator
disguised by a faux wheel cover.

The Brazilian T2c put a new nose on the Bus and gave it a proper radiator grille, but it is prohibitively expensive in the USA to lay hands on all of the Brazilian components needed to convert an original German bus. That and once you start carving up the front of the vehicle, there is no going back. You're either going to finish, or pictures of your vehicle are going to show up in the 'backyard abortions' gallery on the air-cooled VWs websites.

At this rate, its unclear what the vehicle is,
so much of the front is covered.

One of the most consistent difficulties in adapting a radiator to a T2b is trying to produce a solution that work consistently below freezing as well as in Death Valley that is not, frankly, butt-ugly. Mile after weary mile when you're in the middle of nowhere, or every day around the city from stoplight to stoplight. Many cooling solutions will work, in a certain confined seasons or circumstances. But a proper solution must work in all of them and have enough spare capacity built into it that there can be a modest problem and you can still limp to the next village for help.

The first solution is to put the radiator up front, bang slap against the airflow, just the way they did it in Brazil, but without going to all of the panel beating required to graft it in. Unfortunately, radiators hung off the front look like...radiators hung off the front. They ruin the classic lines of the vehicle and no matter how well decorated, (and how wonderfully functional the radiator is in cooling the engine) they still look like a hack job. Those who have gone to the trouble to source the radiator fascia from Brazil certainly have a better looking product. But this gets very expensive, very fast unless you have considerable fabrication skills to also do the body work required to graft in the grille.

Not only does it look like they left
the rear engine hatch somewhere back on the motorway,
also note the size of the extensions on the
 intake vents at the top rear.

Others have attempted to replace the rear engine door with a radiator...and it looks like it. This is based on the idea that the small air intakes on the top left and right rear of the vehicle are sufficiently large to pressurize air through the engine compartment to drive this back-door radiator. This is patently ridiculous: such a design would need to have the space between the engine and the body sealed as the aircooled engine was, to force all of the incoming air through the radiator. The air will always take the path of least resistance if given the opportunity.

Not only do you have a scalding hot radiator out where it can burn you and others, how are you going to get TO your engine in an emergency if the principle access is over 200*F? So this one loses for practicality, aesthetics and cooling.

Engine shown with radiator partly covering
 the battery at the rear, at the bottom of the vent stack.
In this installation, there are small radiators
 on both sides which have been plumbed together in series.

The next to last solution starts to get the idea nailed down: move the radiator inside the engine compartment. Usually two radiators, one at the bottom of each intake vent. With fans, the radiators breathe in from the vents, and out to...where?

Well, down. Down around the space between the body of the engine compartment, and the engine itself. This is probably a workable solution...as long as you keep moving. But when you hit stop and go traffic, or are creeping past an accident in only one lane on the freeway, you won't have enough ram air to move heat from the radiator. So your fan comes on...do we overheat? Maybe.

One successful variant of this design is to seal the radiators to the area around the bottom of the vent stacks. That way all air that passes down the vent stack cannot 'go around' the radiators; it must pass through them. This would dedicate the vent stacks to the radiators, and translate any forward movement to pressurizing the air through the radiator.

But it turns out that performing this kind of seal up is really, REALLY hard. The radiators must fit precisely, and where they don't, they need to have cladding added so that they are both mechanically secure and reasonably airtight. Making all of this ready to go bouncing and rattling down a dirt road is a dubious prospect because of a few fundamental problems remain:

• There is an enormous amount of fitting, cutting, and perhaps welding required to seal both vent stacks.
• The vent stack on the right cannot be permanently sealed, or the battery (which lives there) must be relocated, possibly inside the vehicle under the rear seat, or possibly suspended from the roof of the engine compartment. Regardless, the main power harness would have to be modified or extended to the new location.
• Heat rises. In the air cooled engine, the sealed engine compartment means that hot air coming off of a stopped engine rises up and out of the vents, drawing cool air in from beneath the vehicle, pulling it, in reverse, through all of the fins on the heads and cylinders. But a stopped liquid cooled engine would be trying to use its fans to cool the engine by pulling air down the vent stacks, through the radiators and then down (again) to be pushed out from under the vehicle. This is fighting the heated air's natural tendency to rise and draw in cool air from beneath it. Perhaps not a problem at speed. But three hours of stop-and-go traffic through downtown Atlanta when it is 110 in the shade....your fans would need to be running almost constantly.
• You had not only best have an almost hermetically sealed air intake, you'd better have very very good radiators that can move a LOT of heat. This means ducting and really stellar fans.

There are two enormous benefits to this arrangement, however:

Radiators are kept in engine country instead of body country (the front, the rear, the top) or chassis country (beneath.) Radiators are fragile, and don't react well to rocks, fallen branches, road kill, and other sundries crashing into them at speed. They are well protected, unlike ANY OTHER SOLUTION.

The engine compartment solution answers the 'ugly' problem, and we've seen some pretty bug-ugly example above.

Keeping our cool is the challenge there.

So to sum up, the three radiator placement solutions proposed so far are:

At the front, which is either expensive in terms of body work (if you fit a Brazilian grille) or ugly (and dangerous) if you just mount the radiator on the front bumper.

At the rear, replacing the rear engine hatch. Ugly, also dangerous, and ultimately ineffective because of insufficient throughput.

In the air intake stacks. Only effective with great gobs of fabrication effort to ensure a good seal of the radiator bodies. And it requires either access provisions for the battery, or relocation of the battery. Lots of work, and great reward if done right. If done wrong...baked engine.

There is only one universal place left: beneath the vehicle.

The State of the Art

There have been plenty of attempts to transplant engines in VWs. I'll skip examples that don't (or for practical reasons cannot or should not) apply to the Bus. But it took one of the consistently worst engines Volkswagen ever produced to make the problem acute enough to require serious consideration of anyone wishing to stay on the road: The T3 (Vanagon) Wasserboxer. The "WBX" has a fundamental problem that anyone familiar with an early Subaru EJ25 engine is familiar with: they blow head-gaskets on such a regular basis that you can set your watch by them. Roughly every 30,000 miles. POOF! And your T3 Vanagon is laid up with a horribly expensive repair...again. The WBX was such a turkey that as the T3 continued production in South Africa from 1992-2002, the engine was replaced with a 5 speed Audi unit and finally allowed the total design to show itself as a rugged offroader.

The muscled up T3 with 4WD 'Synchro' and a Subaru power plant.

In the early 1990s, after VW had discontinued the Vanagon in the United States, machinist Hobart Kennedy in Palmdale, California worked out a machined conversion plate and custom flywheel to allow the use of a 90-94 Subaru boxer engine (of about the same shape, size and weight as the WBX) in a Vanagon. He even went the extra mile (several million, according to him) to get the conversion approved by the California Air Resource Board (CARB) which maintains the strictest emissions regime in the world. (MOT & TUV are pushovers by comparison.) To date, Executive Order # D-428-2 is the only approved substitution for the pesky WBX in California.

The Subaru EJ22 - 1990-1996 US Market.
SOHC 2.2L and twice the HP and Torque of the stock
air-cooled engine. Ludicrously overbuilt.

Of course, to make the conversion work, you have more than just a steel plate and a flywheel to contend with. The ECU (Electronic Control Unit) and the entire engine wiring harness from the donor car must come with the engine. Custom exhaust, custom engine carrier, custom air intake... all of the bits coming to about $1700 USD, plus the cost of the donor vehicle from which the engine, wiring and ECU were to be harvested. Pay someone else to bolt it all together and chase bugs out of it, and you're at about $5000. But if you're in California, this is your only option if you intend to stay legal in a Vanagon.

Other pre-engineered solutions have shown up since then: Bostig Engineering and the Ford Zetec 'World Engine.' Inline 4 cylinder VW engines, including Turbo Diesels from late model VWs. Several companies (Rocky Mountain Westy, Small Car, Vanaru, Subagon) have sprung up in the United States and abroad to serve this market, both providing kits as well as making full service conversions for buyers outside the CARB administration zone, but all of them focus on the Vanagon as their target vehicle.

Designed in the mid-1960s to overcome all of the short-coming of the
Beetle engine, the Type4 engine was used in four models of Volkswagen
(411/412 from which it draw its name) as well as two models of Porsche
(the 914 and the one year 912E.) Note that the engine packages
of the Type4 and the EJ22 are almost identical in size.

The only way that this near miss helps me and my 1977 Bus (hereafter, T2b) is the hereditary nature of VW's designs.

Unlike other manufacturers who sell a completely different product line in the developing world than they do in the USA and Europe, Volkswagen is famous for keeping an older design of vehicle in production in a developing market, rather than create a new design out of whole cloth, just for that market. The aircooled Beetle stayed on in Mexico as a taxi until 2003. The VW Bus still lives in Brazil as the T2c, a water-cooled variant of my T2b that will go out of production in 2013.

I'm helped by all of the Vanagon conversions because the Vanagon and the T2b share the same basic transmission: the 091. While gear ratios differ because of where the different engines make their peak power, the basic transmission is the same. So an engine that can be made to mate to a Vanagon 091 transmission can just as easily bolt on to a T2b 091.

There are a few catches...

The T2c, as manufactured in Brazil from 1995-2013.
Why not buy one? Not legal to import in the United States.

1. Just because an engine CAN mate to an 091, doesn't mean that it will fit in the engine compartment. (Some conversions make a false deck so they can raise the roof on the engine compartment.)
2. An engine mated to an 091 in a T2b still needs a carrier or 'engine mount' to support the weight of the engine itself: The stock Type 4 engine carrier bar isn't compatible 
3. An engine rotated or tipped, etc so it will fit in the engine compartment AND align with an engine mount may not be happy being rotated out of its designed alignment. Components like the oil pump pickup from the oil sump rely on gravity to be straight down relative to the engine's intended orientation. Tip an engine too far to one side to make it fit...and suddenly you're starving it for oil at freeway speeds.
4. Even if we succeed in finding an engine that meets, or can be made to meet, all of these requirements, that doesn't mean that ancillaries like alternators, intake manifolds, etc. will fit once bolted into place.
5. We still have to find a way to add components like the non-existent radiator, associated plumbing, and components like an expansion tank and an overflow tank.

Next time, I'll talk about what provisions can be made for these issues.

Every man did what was right in his own eyes....

The Type2b 'Transporter' and period 70's hottie.

If you don't know your Bible, you'd think the title of this post was written about vintage Volkswagen owners. The beloved and benighted vehicles have been subjected to more hacks, shade-tree mechanicals, and jerry-built alterations imaginable. There are a couple of reasons for this:

1. They're historically cheap to buy
2. They're of a very, very simple design
3. Their simplicity belies the level of engineering that went into them. And every doofus with a wrench thinks he can improve on it.

But when it comes to conversions, well...take the normal VW owner's blind tinkering and turn the damnfoolishness up to 11. That's not to say that there aren't some beautiful conversions out there. There are. And the engineering challenge of stuffing an engine into a vehicle that was never remotely designed for it is appealing to people who spent their childhood taking things apart to see how they worked (regardless of whether they could get them to work after they were put back together.) But adapting a diesel engine into what was a gasoline powered vehicle is easy by comparison. We're talking about vehicles which have no place to put a radiator, a rather important lack.

Some of the 'conversion-perversions' are done as entertainment while consuming substantial amounts of beer. Some are 'drive-thru' conversions, where the vehicle is taken to a custom shop and money is sprayed at it until it runs. And some are intellectual exercises, where time, expense, and skull-sweat are the goal and the conversion of the vehicle is just the handy excuse to stave off boredom.

My goal is 'none of the above.' I have an engineer's mind without a mechanical engineer's training. My target vehicle is a 1977 7 passenger Transporter (aka Bus-bus, aka 'Ferdinand the Bulli') that has sat in a downstate Delaware barn for a dozen years before being purchased and relocated to South Central NJ. The vehicle was originally a Champagne Edition 1 which included some uprated upholstery as well as a gas fired cab heater and an uprated alternator to drive it. (The alterations to the body to support these features may eventually come in handy.)

I have almost a quarter century of experience with VWs, having owned most of the air-cooled models, installed and rebuilt my own engines, and having built a show winning 'reference restoration' 1972 VW Super Beetle (1302 model.) I've driven cross country three times in VWs, including solo jaunts in a 1971 8 passenger bus. Cars and I go way back. So when I sold the Super Beetle to a collector after its restoration was complete, I decided that a Bus for the family would be enjoyable. And certainly roomier than Beetle, which proved to be too cramped for my children. 

The daughter examines the bus, fresh from the barn. 

I spent from October 2010 to January 2011 studying the state of parts availability for the VW Bus in the USA. It isn't nearly as rich as it was 5 years ago. There are now only 3-5 engine builders competent to rebuild the Type4 engine used in the vehicles from 1972-1979. Because of the limited manufacturing run of the engine, consumables (like exhaust, emissions components, Fuel Injection components) are becoming  scarce or are available only as rebuilds, with a fraction of the lifespan of the original.There are some attempts by enthusiasts to do limited manufacturing runs of some components to keep their rig alive, and I fully support them. These enthusiasts are doing it 'for the love,' which is good, because there is no faster way to lose your shirt than to labor under the delusion that VW owners are anything but cheapskates and no matter how clever your 'replacement parts' fabrication is, you're likely going to lose money on the deal.

That said, there is a breed of cat that will pay a reasonable amount and then chip in his own sweat equity to the cost of the product. That's the rational DIY guy. That's me. In my day job, I constantly have to weigh compromises between 'build,' 'buy,' and 'outsource' (e.g. pay someone else to build something custom just for my company.') This is all in the hazy realm of software and hardware for high performance computing. And the triangle at the right is the balancing act that must always occur. Note that at no point do all of the Venn Diagram areas overlap each other. One of the three is always compromised in achieving the other two.

So I'm going to take what I know about the 'iron triangle' of project management, add what I know about VWs, and then attempt to work the project of putting a Subaru power plant into my VW bay window Bus. Not just because I want a solution that will insulate me from constant fiddling with the engine to keep it running or inspected, but as a 'worked example' for others to follow. Partly 

There have been some great examples by people intending to do the same thing: show their work. The difference with this is that I'm going to document every decision along the way showing how I arrived at my conclusions, why I think they're the best solution for the general user who wants to DRIVE their beloved VW, and then how I accomplish it using the experience and knowledge of others (outsource), my own problem solving (build), and the engineering that others have already done (buy.) I haven't set a budget yet, but I intend to soon. My intent is to pull the complete conversion in BELOW the cost of a single reputable Type4 engine rebuild (about $3500 and good for about 80-100k miles.)

As a final pièce de résistance, I intend to perform this conversion without carving on the vehicle in casual or non-reversible ways. My goal is to made more difficult by a commitment to myself that the vehicle be able to return to a factory engine configuration without having to reconstruct the wiring harness or back half of the vehicle. Old engine out, any connectivity to the existing body, transmission, electrical, etc buffered by a breakout box so that the old engine and original components can slide back in. No permanent changes. 

Vorwärts!