AutoSpeed - Project Honda Insight, Part 5

This series is based around a 2001 model hybrid Honda Insight.

The Insight remains one of the most aerodynamic and lightest cars ever made, with a Cd of 0.25 and a total mass of about 850kg from its 2-seater aluminium body.

The intent of the project is to turbocharge the engine, add water/air intercooling and programmable engine management, and then provide new high voltage batteries and a new electric motor control system.

The aim is to build a car with the best performance/economy compromise of any in the world.

The series so far:

Project Honda Insight, Part 1 – Introduction

Project Honda Insight, Part 2 – Fitting an Alternator

Project Honda Insight, Part 3 – Building an Airbox

Project Honda Insight, Part 4 – Intercooling Requirements

This issue: building the water/air intercooling system.

The intercooling system in the Honda needs to be able to vary intake air temp at will. In this story we make it happen – and then find some major problems in the design.

Recap

Last issue in Project Honda Insight, Part 4 I introduced the water/air intercooling system being used in the Honda Insight.

To recap the system requirements:

The ability to vary intercooler efficiency as required. This is so that warm intake air can be provided when it is needed for best fuel economy. Conversely, when running higher boost, the air can be cooled by the system. This variation in efficiency will be achieved by controlling water pump speed.

A very high intake airflow capacity, so resulting in a low pressure drop. This is achieved by oversizing the components – as an educated guess, the system as fitted to the Honda would be effective on a car developing two to three times as much power.

The intercooling system had to be compact enough to fit in the Honda’s small engine bay, and had to meet reasonable cost requirements.

The recap the system components:

Underbonnet heat exchanger – commercially available, aluminium, universal liquid/water to air intercooler. This uses 63mm (2.5 inch) inlets and outlets with half-inch NPT water fittings. The overall size of the unit is 230 x 230 x 90mm.

Front-mount radiator – small aluminium air/air intercooler core adapted with water fittings. It has a core size of 300 x 155 x 64mm.

Pump - Johnson Controls CO30P5-1 with a spec’d flow of 23.3 litres a minute at a head of 15kPa.

Header tank – aluminium tank of 750ml capacity.

Installation

Heat exchanger

The first step was to lose the 12V battery from under the bonnet. The battery is usually mounted near the firewall on the left-hand side (all locations as viewed from the driver’s seat). The battery is mounted in a deep alloy tray suspended from a transverse strut brace working with an additional dogleg tube.

Removing the battery, and subsequently the dogleg aluminium tube support, gave space above and behind the gearbox for the heat exchanger. The transverse strut brace remains, with the dogleg removed and the resulting hole in the brace TIG welded and ground back. (Subsequently, it was replaced with a mild steel square tube.)

A supporting frame for the heat exchanger was made from 3mm steel sheet and 30 x 4mm steel bar. This frame bolts to the gearbox and an engine mount, and locates the intercooler horizontally but slightly rotated from being ‘square’ to the transverse axis of the car.

Because this frame is positioned under the heat exchanger, and because each of its lower mounting lugs is at a different height, multiple templates and dummies were made up before the final mount was welded-up.

Because the heat exchanger is subject to engine vibration, rubber mounts were interposed between the mounting frame and the intercooler. Folded aluminium brackets attach the heat exchanger to the rubber mounts. These brackets bolt to the welded-in blocks that are located on the end tanks of the heat exchanger.

Note that on my heat exchanger as supplied, the faces of the mounting blocks were not flat (so meaning that without remedial work, the brackets wouldn’t stay tight), and the threaded holes are not located in the middle of the blocks.

Header tank

The aluminium header tank was made from an old aluminium fire extinguisher. An aluminium radiator filler fitting was TIG welded to the top, a mounting bracket added, and a ¾ inch NPT female aluminium fitting welded to the wall near the base. A standard radiator cap is used.

The header tank was mounted on one of the heat exchanger supporting brackets. The heat exchanger bracket was modified with an extension that mated to the mounting bracket welded to the tank. Mounting the tank in this way means it is supported on the same rubber mounts as the heat exchanger.

The intercooler heat exchanger, header tank and supporting frame were all blasted and then powder-coated in wrinkle black. The steel supporting frame underwent another step: after blasting, it was zinc undercoated before being powder-coated.

Front mount radiator

The front mount radiator uses a small air/air intercooler. It has a core size that is 300 x 155 x 64mm and weighs about 1.8kg (empty).

I used my lathe to turn-up some aluminium hose fittings that I then TIG welded inside the standard intercooler hose fittings.

This resulted in a compact, excellent flowing radiator that contains 1 litre of water. The radiator was etch primed and then painted black.

The radiator is fully rubber mounted. Note that these lower mounts are shaped in this (slightly odd) way to provide clearance for the wrap-around lower edge of the bumper cover – only a small amount of plastic needed to be shaved from the inner grille insert to provide clearance.

Pump

The pump is located ahead of the front left wheel. Standard in this location is a resonant chamber tee’d into the intake system prior to the standard airbox. With the revised intake, this resonant box was no longer needed.

So that different pump mounting arrangements could be trialled for their quietness, the pump was initially operated from an external power supply while circulating water into a bucket. This showed that while the pump is relatively quiet, at full battery voltage and under load, ‘hard’ mounting of the pump transmitted an unacceptable amount of noise to the body.

To minimise this noise transmission, the final mount uses a saddle clamp that surrounds the rubber-wrapped pump motor. In turn, the clamp bolts to short rubber mounts attached to a steel mounting bracket. This ‘soft’ mounting of the motor means that very little vibration is transmitted to the aluminium bodywork of the car.

The hose runs to and from the pump are ¾ inch (19mm) heater hose. To prevent movement of the car engine being transmitted by the hoses to the pump (so potentially fatiguing the plastic pump barbs), the hoses are clamped to the bodywork prior to the pump.

Note that the pump must be located so that the pump head is below the pump motor – this is so if any water leaks occur within the pump, water drains away from the electric motor, rather than into it.

Water plumbing

Getting the water plumbing right was surprisingly difficult.

The water fittings supplied with the heat exchanger are straight, barbed ¾ inch hose fittings. However, the straight fittings were not suitable - to provide sufficient clearance, right-angle fittings needed to be used at these two locations.

The endplates of the heat exchanger use female, half-inch NPT threads. NPT (National Pipe Thread) is a standard US tapered thread for pipes and fittings. Unfortunately, here in Australia, the standard for fittings is not NPT – it’s BSP, or British Standard Pipe.

Sourcing right-angle fittings for the heat exchanger was therefore difficult – I simply couldn’t find right-angle, male, NPT, half-inch-to-¾-inch barbed hose fittings at a reasonable price. In the end, I sourced male/female, half-inch NPT right-angle adaptors into which the barbed ¾ inch barbed hose fittings that came with the heat exchanger could be screwed.

However, there was a problem with these right-angle fittings. Internally, they necked-down in size a lot – it appears that the internal hole on one of the two angles has been formed at much less than full size. Since these two right-angle fittings are likely to be among the most restrictive elements in the water flow path, I milled-out the interior of the fittings to give much better flow. (The top fitting in this pic has been modified.)

The hose runs were laid out in ¾ inch (19mm) heater hose. Sufficiently gentle bends were able to be used that the ostensibly straight hose could be persuaded to follow the required paths – except at three points that needed pre-formed bends.

As the pump needed to be orientated so that the pump head was at the bottom, the inlet hose to the pump projected downwards. A tight U-shaped hose was therefore needed to connect to this. A pre-formed heater hose – a CH1963 hose to suit a Ford - was used for this purpose.

Another CH1963 was used to provide the right-angled bend needed at the base of the water reservoir. A hose with pre-formed bends was also needed under the heat exchanger.

The hose runs were clamped in place wherever they could potentially rub against other components. The clamps used were of the ‘P’ design, coated in rubber and having a nominal hose diameter of 20mm.

Testing – and finding problems!

With the system installed, it was time to add coolant and test for leaks. However, just the first step – filling the system with coolant – proved to be problematic.

The issue was that despite the system having a theoretical water volume of about 3.3 litres, getting more than 2.5 litres of coolant into it was impossible. So what was going on? The problem was that air remained in the system.

As shown here, the water fittings on the heat exchanger are located half way up the end blocks. Fill the system with water, and the top half of the intercooler heat exchanger remains full of air, not water! (Note that any water/air intercooling system using this commercial heat exchanger in a horizontal position will have the same problem.)

And I made exactly the same mistake with the design of the front-mount radiator (the modified air/air intercooler). Because the water connections are half way up the end tanks, the top half of the radiator will always be full of air.

Therefore, as installed in the car, the system could not be fully filled with water!

And if this issue wasn’t already serious enough, another problem was then found.

Irrespective of how the water plumbing was arranged, the Johnson Controls pump cavitated badly.

Pump cavitation occurs in an impeller pump as a result of a drop in pressure of the liquid moving through the impeller’s eye. This reduced pressure causes bubbles to form and then quickly collapse. Cavitation can be clearly heard – the pump sounds ‘bubbly’, rather than humming smoothly. In this case, cavitation appeared to be occurring because not enough water could get to the pump – that’s despite it being located at the lowest part of the system.

Indicating that cavitation was the problem, the noise stopped when the pump was run at slower speeds – but at 13V, cavitation was intense and continuous.

Cavitation had not occurred when the system was tested on the workshop floor, but with the tighter hose bends present in the engine bay, pump cavitating noise could be heard and air bubbles could be felt passing along the hoses. Not only would this cavitation reduce the efficiency of the system, but it would also cause the pump to wear prematurely.

With a water/air intercooling system unable to be filled with water, and a pump unable to effectively circulate the water in the system, a major rethink was needed.

Project Honda Insight, Part 5 – Intercooling System #1

Building the intercooler system

by Julian Edgar

Recap

Installation

Testing – and finding problems!