Last week in Technokill: Building a Blown Hybrid, Part 1, we backgrounded what needed to be considered before fitting a supercharger to a hybrid petrol/electric car – a first model Toyota Prius. This week we get on and do it!
Fitting the Blower
While on paper fitting the small ex-Subaru Vivio AMR300 supercharger looked easy, in practice it was a nightmare job.
This was primarily because of the very cramped engine bay in the Prius, which necessitated (for example) integrating the supercharger mount into a completely new fabricated engine mount. This bracket was made from mild steel, using 9mm plate, 12mm plate and 40 x 8mm bar, arc welded together. In addition to the heavily gusseted engine mount, the bracket also incorporated the fabricated belt tensioner. In all, something like 40 hours of work went into making the bracket.
When spending the many hours working on the car, it was always in the back of my mind that the supercharger may prove a complete failure – something which isn’t conducive to lots of effort! In addition to the points made last week, there were further concerns:
So rather than continue working until the system was completely finished – only possibly to find it could never work – I therefore resolved to do some testing with the system only half-installed. So with the supercharger running (but drawing air through a hastily put together and fairly restrictive intake, and with the intercooler projecting out the front of the car, held in place only with hoses and clamps!) I undertook a very careful test drive. The bumper wasn’t fitted, the bonnet was held down with wire, and the numberplate was cable-tied into place. Mixtures were monitored with a MoTeC air/fuel ratio meter.
Because no supercharger bypass had yet been fitted, and because the supercharger was blowing into the throttle, at idle the intake air rapidly became hot. This lack of recirc valve also meant that the throttle had to be closed very slowly, otherwise the pressure build-up caused a hose to pop off.
However, the results of this preliminary test were a huge relief. The boost level was between 5-7 psi, the car didn’t detonate, the air/fuel ratios were satisfactory (well within the range of being able to be tuned with the Digital Fuel Adjuster), the multi-rib belt drive (although having more whip than wanted) didn’t come off and there was no apparent belt slippage, and the intake air temps after the intercooler were low.
So I could continue with the installation, which by this stage looked like it would take up a full month of work.
Installing the Intercooler and Airbox
The next steps were to properly install the intercooler and airbox, and make the plumbing.
The intercooler – from a diesel Mitsubishi Pajero – was to be mounted at the front of the car. It could be placed either behind or in front of the power converter radiator. (The underbonnet power converter runs its own dedicated water cooling system, complete with pump and front-mounted radiator.) Obviously, from a point of view of maximum cooling, the intercooler would be best at the very front of the car. However, autos.groups.yahoo.com sources suggested that during development of the Prius, Toyota engineers had difficulty keeping the temperature of the power converter down – and so the radiator for this dedicated system shouldn’t be blocked.
But then again, keeping the boosted intake air as low in temp as possible would also be very important!
The decision was made – mount the intercooler in front of all the other heat exchangers. In this position it blocks about half the area of the power converter radiator.
In order that the intercooler could be mounted in this position, the power converter radiator needed to be moved backwards and the frontal intrusion bar modified to provide clearance. I cut out a section of the original bar and then had welded across the gap a section of the bar from the half-cut Prius that I had bought. Note that this bar is slightly curved in plan, so it couldn’t just be replaced with straight rectangular tube.
The standard airbox sits on top of the engine and its outlet is in the wrong place to feed the supercharger. A new airbox was sourced from a Daihatsu Sirion and was mounted in the front guard. The original Prius airflow meter is built into the airbox. The airflow meter was carefully cut out so that it could be mounted conventionally in-line after the new airbox. The wiring loom to the airflow meter was extended to reach the new airflow meter location.
The supercharger inlet, outlet and intercooler plumbing was fabricated primarily from 50mm copper tube and fittings which were brazed together. Taking this approach gave high-flow bends which were still tight in radius, and also allowed much of the tubing to be fitted together (the bends incorporate sockets into which the tube pushes) before being taken to the welder. However, after the plumbing was in place, a major problem appeared. The tubes running to the intercooler passed over the top of the radiator, before sharply bending downwards to reach the intercooler.
And when these were in place, the bonnet wouldn’t shut....
While checking clearance to the bonnet seems an obvious step, in the case of the Prius the very steeply sloping bonnet (which also has an underside that actually wraps over and around the bonnet locking platform) meant that clearances in this area were far tighter than in other cars. To overcome the problem, the front tubes needed to be changed from copper to rubber, and part of the underside metalwork of the bonnet had to be (very carefully!) cut away.
There were also two other very tight spots – the inlet and outlet of the supercharger. The inlet, closest to the back of the car, is located very near an intake manifold head stud and the variable cam timing solenoid. The copper inlet pipe had to be hand-beaten to shape to clear both of these obstacles. The outlet was similar, although in this case it was the radiator filler neck and radiator fan shroud to which clearance was required. In fact, this area was so tight that the radiator was pulled forward a little at the top. As a result, both the inlet and outlet connections to the supercharger are smaller in cross-sectional area than would be expected with 50mm tube; however the reductions in cross-sectional area are made smoothly and so the flow will not be as restricted as it would if these changes were abrupt.
The placement of the intercooler plumbing and the moving forward of the top of the radiator meant that the original panel across the top of the radiator had to be replaced with a new panel. This was made from aluminium channel. Foam rubber was used to block the gap by which air could have otherwise bypassed the radiator. The removal of the panel above the radiator also meant that the bonnet lock needed to be supported by other means; an aluminium bar was bolted across the car in front of the air-con condenser and this carries the bonnet lock and the horns.
A supercharger bypass valve was incorporated in the system. This comprises a GFB recirculating blow-off valve, normally used on a turbo car. It has 32mm plumbing and is beautifully made. The connections to this valve were plumbed-in just after the airflow meter and just before the throttle body.
It was literally only at this stage – after nearly 3 weeks of work – that the system even started to look near complete.
In order that – if necessary - changes could be made to the supercharger bracket, the inlet and outlet plumbing, the airbox and everything else new, the second test drive was undertaken without any of the supercharger system components painted.
When the engine was started, the supercharger was very loud – because of insufficient intake manifold vacuum, the bypass valve wasn’t opening. I disassembled the GFB valve and shortened the spring, which reduced the preload and allowed the valve to open at idle. However, the supercharger noise through the new airbox was still fearsome – far louder than desired. Blocking off some of the air inlet to the box decreased the noise substantially, indicating that the noise could possibly be reduced through further intake tuning.
However, the next problem was much more major: driving the car showed only 2 psi maximum boost. After checking clamps and blocking off the blow-off valve in case it was leaking, a simple mistake was found – the crankcase breather hose was off the intake manifold. With this huge boost leak plugged, boost again rose back to the previous level.
The belt drive had settled down (I gave it greater tension that stopped its whipping) and on the road the car was performing fine. The mixtures were a little leaner than preferable, but that could be easily changed with the Digital Fuel Adjuster.
So at this stage things were looking very good - but they weren’t to continue that way....
Next week: on the road with the world’s only supercharged, intercooler battery/electric hybrid