Project Honda Insight, Part 7 - Turbocharging

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 Project Honda Insight, Part 5 – Intercooling System #1 Project Honda Insight, Part 6 – Intercooling System #2 This issue: turbocharging the engine

The turbo

The selected turbo for the Insight is a Mitsubishi TF035HM-13T-6 (also coded 49135-05000). It is usually fitted to a Fiat / Iveco Daily II commercial vehicle, a 2.8 litre diesel that develops 90kW and 290Nm. In that engine, peak power is at 3600 rpm, peak torque at 1800 rpm and boost is quoted at being 1.4 Bar. The engine is a four cylinder, 8-valve design.

So why on earth would I select a turbo from a 2.8 litre diesel engine for a 1 litre petrol engine?

The primary basis of selection was simply physical size – this is a small turbo that to my eyes, looks about right for a 1-litre engine. I’ve owned factory 660cc turbo three cylinder cars that boosted really early, and this turbo is just a bit bigger than those turbos.

There’s another way of looking at it too. If we multiply engine capacity by half engine revs (half, because a 4-stroke inhales only once per two revs) and then multiply by an estimated volumetric efficiency, we get a ‘naturally aspirated’ airflow number.

So, for the Fiat / Iveco, that would be 2.8 litres x 1800 rpm x (say) 70 per cent volumetric efficiency = 3528 litres/minute.

Now we do the same for the Honda. The Honda breathes well (peak torque is at 4800 rpm) so we’ll assume a volumetric efficiency of 100 per cent at those revs (and it may well be higher). So 1 litre x 2850 rpm x 100 per cent = 2850 litres/minute.

Do these sort of back-of-envelope calculations shows the 2.8 diesel to be flowing about 24 per cent more than the Honda – without boost. So despite the diesel being 2.8 times the volume of the Honda, the likely airflows are much closer.

But there’s another factor. The Fiat / Iveco runs a boost pressure of 1.4 Bar (about 20 psi) – far more than I intend running in the Honda. In the Fiat / Iveco this boost pressure is developed by 1800 rpm (that’s where peak torque is, remember). So if this turbo can develop 20 psi at 1800 rpm in the Fiat / Iveco, will it be able to develop (say) 7 psi at 2000 rpm in the Honda?

My guess is ‘yes’ – but it is only a guess.

I hope that the guess is correct because for the turbo to be really effective in the Honda, full boost pressure will need to be develop by about 2500 rpm at full load. That’s because with the very high gearing of the car, the engine is ticking over at about 2500 rpm at highway speeds like 110 km/h.

Of course, there’s another puzzle in the turbo matching. When the car gets back its 10kW (or more!) of electric assist (a second stage in the modification process), the bottom-end torque production of the driveline will be completely changed….

The selected turbo is therefore a starting point - if it works well, great. And if it doesn’t, changing it for a turbo of a slightly different size will not be a huge job.

The Mitsubishi turbo is brand new – it was picked up on eBay some years ago.

Turbo mounting

One benefit of turbocharging the Insight is that no exhaust manifold is needed. This is because the head design incorporates the three-cylinder’s exhaust manifold within the head casting. The standard exhaust outlet is a single, oval-shaped hole. In the standard car, the exhaust pipe bolts straight to this hole using a three-bolt flange.

However, the exhaust housing of the Mitsubishi TF035HM-13T-6 turbo uses a relatively small round entrance hole.

To connect the Honda’s head to the turbo therefore requires an adaptor that changes from an oval cross-section to a round cross-section of significantly smaller area.

To achieve this adaptor I used a steam pipe reducer, 2 inches (50mm) at one end and 1.25 inches (32mm) at the other end. The schedule 40 adaptor is made from stainless steel and uses a thick wall.

Starting with the large diameter end, I squashed the adaptor in a hydraulic press until it conformed to the oval shape required at the ‘engine head’ end of the adaptor. To prevent the other end from being crushed at the same time, I inserted an appropriate mandrel (an old socket) into it during the pressing.

The other end of the cone-shaped adaptor was then cut at the point where its internal diameter matched the turbo entrance hole diameter.

A new flange was then made for the turbo (I used the original exhaust flange at the other end) and the adaptor MIG-welded to the flanges. Care was taken that the internal transition from oval to round was smooth. The result is an adaptor that should rapidly increase the speed of exhaust gases passing into the turbo’s exhaust housing.

Turbo changes

Two major changes were made to the standard Iveco turbo configuration.

Firstly, the compressor and exhaust housings were rotated to provide the best configuration for the Honda installation. In the Honda, the compressor outlet needs to point vertically, and the exhaust inlet needs to be near-horizontal and pointing forwards.

Secondly, a new wastegate actuator was fitted. The standard Iveco wastegate is set to around 20 psi – over double what I need. A 4 psi wastegate actuator was fitted, with the compressor housing needing to be redrilled to mount it in the correct position. (Boost higher than 4 psi will be able to be run by using a wastegate control valve.)

Turbo oil supply

The turbo oil supply is picked-up from the spot on the front of the block that originally housed an oil pressure switch. The requirement was for not only a supply of oil for the turbo but also a mounting point for an oil pressure sensor to replace the original switch.

Making a hose to connect the oil pressure port to the turbo can be a nightmare - what with different sized and threaded fittings, the requirement for an extra port for the sensor, and so on. I therefore got a specialised hydraulics supplier to make up a custom hose.

I took to the supplier the original oil pressure switch (this would then show the block port’s hole diameter and thread), the new pressure sensor, the turbo, and a piece of bent wire showing the length and path that the hose would need to take. Prior to going to the supplier, I also measured the clearance available for the adaptor and the new oil pressure sensor.

With all of these bits available, the supplier was able to make up a system to suit. It comprised a steel adaptor that screwed into the block, a re-tapped hole in the adapter to suit the pressure sensor thread, a right-angle fitting to which a crimped braided hose fitting attached, the hose, and then a banjo fitting that screwed into the oil supply port on the turbo.

Turbo oil drain

To install the turbo oil drain, the magnesium alloy sump was removed and drilled for a steel bulkhead fitting. This fitting in turn connected to a right-angle brass fitting and then a brass hose barb.

Steel washers were used either side of the sump to allow the sealant to have a much greater area of contact. Loctite was also used on the inner nut.

Oil temperature sensor It’s not strictly related to the story of installing the turbo, but it was carried out while the sump was off the engine. What’s being described? The installation of an oil temperature sensor. To form a suitable mount, a standard steel hydraulic bulkhead fitting was selected. The surplus threaded end was cut off with a hacksaw and the internal hole tapped with a thread to suit the temperature sensor. The fitting was installed, using steel washers either side of the sump wall. The sensor was then screwed into place. The interior view.

Heat shield

A heat shield was built to prevent heat from the turbo’s exhaust housing cooking nearby engine bay parts. Commercially available ‘sandwich’ heat shield material was used – this is able to be bent, cut and shaped.

The heat shield material can be joined using nuts and bolts, or as shown here, with pop rivets. The rivets attach to small angle brackets placed on the other side of the sheet.

The sheet was stiffened by the use of rolled edges top and bottom. These were formed in a two-stage process: first the edge was partly folded with angled long nose pliers, then a hammer and curved mandrel was used to give the full 90 degree bend.

To hold the heat shield securely in place, this steel bracket was made. It bolts to the exhaust pipe flange and provides three mounting points (the lower one is not visible here) for the heat shield. Rivnuts (arrowed) were used to facilitate the mounting – there’s therefore no need to access the nuts when screwing the heatshield into place.

The completed heat shield, mounted on the turbo. I chose to go over the top of the wastegate actuator rod – there was plenty of room and it created a bigger air gap. However, the heat shield more closely follows the shape of the lower part of the exhaust housing – here there was less room available.

Exhaust brace

A strong brace (arrowed) connects the exhaust after the turbo to the alternator mount. This brace is used to reduce the loads occurring on the three 8mm studs that hold the turbo exhaust manifold to the head. In effect, the turbo is supported by three head studs, plus the four mounting bolts of the alternator bracket.

Incidentals

- Boost control valve

A Mac BCS 35A-AAA-DDBA-1BA boost control valve was installed (arrow). This 3-port design is a special high temperature version and is compatible with a 10-30Hz Pulse Width Modulation (PWM) frequency. At this stage it’s mounted but is not controlling turbo boost – that function will come with the installation of the MoTeC ECU.

- One-way valves

One-way valves were installed on the brake booster vacuum feed (arrowed) and also the return line for the charcoal canister manifold line. These valves are Mitsubish Magna brake booster valves sourced from a wrecker.

Conclusion

As with most turbo installs, it was all the fiddly bits that took the time – making the heat shield, getting the oil feed and return plumbing organised, and so on. But here’s how it looks when finally installed under the Insight’s tiny bonnet, tucked between the engine and firewall.

Next: building the exhaust

Did you enjoy this article?

Please consider supporting AutoSpeed with a small contribution. More Info...

Share this Article:

Tweet