`Ratting around in the garage 

Note : This is written in New Zealand.
Approx.  NZone$ is worth US70c

For openers

So you want to be a success? Well there are a few general principles that will hold you in good stead for just about anything; work projects, career, home improvements, marriage(?):

I firmly believe (with my hand on my heart) that in building the Road Rat not one of the above has ever been allowed to even poke itís nose through the garage door, (except of course the fine welding by Alan and Phil, top notch machining by Roy, and Class A fibreglass makings from Alex Mac).

This article is a follow-up to the presentation of the car to the club at the April meeting and randomly covers a few of the issues I faced in getting it from light-bulb-in-the-air-above-my-head stage through to street legal. Other places where details of the car are to be found are :


Why build a car anyway? What a dumb question. However, I am very much of the persuasion that if you are going to build a car anyway, then you better make sure that it is not like anything that you could have bought in a shop (or from Ebay / Trademe). Because, lets face it, Toyota, Ford, Porsche and their cousies have put a lot of effort in turning out pretty sound products that do most of what you could ask for pretty well. And they donít really cost all that much. So there is limited satisfaction in investing several thousand hours in making something that you could have bought anyway. There has to be another reason. Why I like building cars is that it gives you a blank canvas (or in my case, more a back of an envelope really) to indulge yourself with any flight of fancy that you might have.

What are we actually talking about here?

The Road Rat is a minimalist space frame car, two seats, no roof, mid engine, built mostly from 60ís and 70ís VW parts. The body is 7ish, with aluminium flat panels on the sides, front cycle guards, fibreglass nose-cone, rear guards and "boot" lid, and a curved textured aluminium rear transom.

Howmush didid coshcha tabildid then?

A very important question. Always comes up early in the roadside chat, just after "What motor have you got in it?". The answer is - probably about $20 per week. This car took me seven years, and this period included a lot of down time, and a lot of other things going on. There were no big expense items but a lot of stuff was bought and not used. The biggest single expenditure was the four alloy wheels at $200 each. There were a lot of things that cost about $100. But basically it is cheaper to build a car than it is to smoke cigarettes.

Body bits

I built the body as a series of separate panels which are fastened on to the frame with stainless steel screws and captive nuts. Similar to a motorcycle, these panels are relatively easy to remove to get access to mechanical things, or to repair or reshape them. I have enjoyed the freedom of form that fibreglass construction offers, especially the way that common household objects can be pressed into service as moulds. For example:

Where possible I have tried to use female moulds for fibreglassing so as to get a smooth finish surface straight off. However, in most cases there had to be so much reshaping afterwards by cutting bits off and glassing them back on that the advantage of doing this was pretty much lost. Most of the body bits were finally moulded directly on the car by arranging the separately made pieces and filling in the gaps with fibreglass. In making a shape for this, I wired a wooden substructure onto the car chassis, glued and/or screwed more wood onto this and then formed the necessary shapes with MDF board, corrugated plastic sheet (garage sale signs) and gib-board plaster. I found that it is not necessary to let the plaster cure fully. It works OK when still a bit damp. For bond release I used mostly packing tape. This is much quicker and cleaner than wax and guaranteed to release easily. For the first "glass-up" I used just one sheet of chopped strand mat, and then added additional layer(s) to the backside later, when I was happy that everything fitted and looked right. Final finishing involved a bit of bog and a lot of sanding. There is much less than one litre of bog in the car, as most of it gets sanded back off. I note that I am talking about body panels with a maximum dimension of less than one metre. This technique would work OK for a single whole-car moulding, but there would need to be extra stiffness built in somehow.

Before going with fibreglass I experimented with beating some aluminium with a view towards making all the bodywork in this material. My thought was to use some beaten and some single curvature sections riveted together like you see on the engine covers of an aeroplane. However I gave up on this pretty quickly. It was too hard. And, I am glad I gave this away as fibreglass panels are so easy to modify and so tough. They are also easy to "recondition" after the inevitable scratches etc. happen.


My word, those proper car-painting businesses do an excellent job. Needless to say I opted to have a go at this myself. The main reason was that I was completing different parts of the car at different times. I would have liked to have a nice powder coated chassis, but there were obviously going to be a number of changes and lots of attachment points to be welded on as the car grew, so to keep rust at bay and avoid the necessity of having to pull everything apart one more time, I tried to keep everything made of steel painted as I went. For the fibreglass bits I thought, well, itís a tiny car so for safety reasons a bright colour is important. From then on the philosophy went something like : We are used to seeing a certain fashionable range of colours on cars. These are chosen by people who do market research and are targeted at making their product appeal to a group of statistically generated non-real punters. There is no need to be part of this process. In fact, it is your duty to clearly not be part of it. So, fluorescent red it was for the nose cone, and finally, a similar green for the rear panels. I did all this with spray-cans; first a coat of white, then a few coasts of the colour, then clear polyurethane. I found that it pays to get all the paint from the same manufacturer. I learned this the hard way when I noticed that The Warehouse paint was causing the previous layer to bubble up. All this hasnít really made a quality paint jobÖ but it looks great at 50 metres and/or 100 km/hr.

Steering : Ackerman humbug

Lets not get too technical about the VW T3 trailing arm front suspension geometry, but with the shortened Toyota steering rack located to achieve negligible bump-steer, the result was negative Ackerman. This means that on full lock, the outside wheel was at a greater turning angle than the inside wheel. Unless you plan a lot of very serious tail-out cornering, this is not right. I was a bit suspicious of how much the theory of Ackerman angles actually is used in cars today so I did a bit of research. On full lock the front wheels of my old Sherpa van were at 30deg and 40deg. That seemed about right. Our BMW was something like 32deg/36deg but the several Japanese cars I checked, even a RWD starlet, were all 30/30 .

So significant Ackerman for light cars seems now to be out of fashion. Still I needed to at least get the Road Rat into positive Ackerman, and this was going to need new steering arms. This was also an opportunity to increase the steering ratio. Nothing feels worse to steer than a sporty car with a slow steering rate. It is also dangerous with a short wheelbase car because when you need opposite lock, you tend to need it real quick. No time for any hand over hand business. I designed the new steering arms so that no cross section on them was less than twice the moment of inertia of the equivalent donor vehicle part. Coupled with the much lighter weight of the new car, this seemed a safe design approach. These were profile cut out of 32mm plate for me, then I drilled, hacksawed, ground and filed then into the final shape. Roy helped put the tapered hole in for the tie-rod end. Toyota, VW, BMC, Honda, Mazda and lots of other cars use the same 14mm thread for tie rods, and VW uses hollow tube tie rods which can easily be cut and treaded to suit the required length. The final outcome was about +2deg of Ackerman at full lock, and 2.25 turns of the steering wheel lock to lock.


Having the right standard numbers on the various lights seems to be really important to the powers that be. The simplest way to address this is to just buy ADR (Australian Design Regs) compliant bits from your local car parts shop. Not wanting any more hassle than necessary, in the end this is what I did. And it worked out very well. The "Trailer" lights on the back seem rather gauche, but they are actually really good. They are cheap, easy to put on, easy to wire up, easy to replace and they are good and bright. In fact, without reversing lights, in the dark I can see quite clearly enough to reverse down our curved driveway and into the garage with just the light of one blinker or the odd touch on the brake pedal. I think this requirement for lights meeting a certain standard is a good thing.


Here I had some trouble. I tried to use all the VW bits in their normal roles. The donor car was a type three VW (Variant) with disks at the front and wider drums at the back than as used for the beetle. I replicated the floor fixing detail for the pedals. The result was very effective brakes, but a spongy feel to the pedal that didnít give confidence. I replaced at lot of parts and used up about $50 in brake fluid with bleeding, but to little avail. I then took it to a local garage where the proprietor, formally a sports car club scrutineer, bleed it again and advised that he would give it a warrant, and that VWís are often a bit like that.

Back to basics. The VW has a mechanical advantage on the pedal of about 11:1 . This is much more than the more modern power brake systems which seem to only have about 5:1 on the pedal. This gives a good feel, but will not give much hydraulic pressure without the servo. So I reckoned that the sponge in the VW system was lack of rigidity in the structure between the pedal and the master cylinder. Then I was watching the Bathurst race on TV and the "pedals cam" showed the driverís feet. The top Holdens and Fords all had bottom hinging pedals, even though the original cars had had top hinging pedals. The way forward was clear. I built a whole new bottom hinging brake pedal box for the car. I made it very rigid and complete with new pedal and a new master cylinder (mail order for USA). I gave it about a 6:1 ratio and bolted it into the car. It works great. Very good feel and can lock the wheels up at 100km/hr without too much effort.


I was not so smart with the wiring. Because the steering column is from a Toyota Starlet and has all the switches etc. on it I thought the best way to do the wiring would be to emulate the Toyota system for the whole car. I had the Toyota fuse box etc, and thought all I would have to do would be to lengthen the wires to the engine a bit. Wrong. I did all this on a big sheet of plywood with wires zigzagging all over it and it all turned out to be very tedious. I tried to reverse-engineer the Toyota wiring, to understand why they had done it the way they did. Then reinterpret this for my purposes. I used up about three wiring looms for extra wire, valiantly persevering with maintaining the original colour coding etc. I stuck to it, and made it all work, in all aspects but the headlights. I just failed to see reason in the way Toyota had done the headlights the way they had, and I put in an extra relay and used my own design for the headlights.

I will pass on the lesson I learned from this experience. Do it this way:

  1. Buy some rolls of new wire, any colour will do.
  2. Design the system you want and understand.
  3. Put all your lights, switches, fuses etc. in place on the car.
  4. Put in the grommets.
  5. Run the wires one at a time. Joint each one up, or at least label it.
  6. Wait until the car is almost complete.
  7. Tape up the loom in place in the car.

Fuel tank, gauge etc.

The tank is from the original VW with a different filler neck soldered on including a provision to vent air from the other side of the tank back to the filler neck. I also fibreglassed on a catch basin and drain round the filler cap to avoid any mess and reduce danger from problems at the filling pump. The tank had to be panel beaten a bit at one corner to give clearance for the right shock. The sender unit is original VW, but the gauge is Citroen as I thought I needed to use a 12v component here. (the Road Rat is 12v although the donor car was 6v). I now got the gauge reading very accurately with a few resistors strategically added to the system. This did not work out so well in the first attempt as I initially used a very low current variable resistor to affect the calibration, and it burned out. This resulted in a false "half full" reading that caused me to run out of fuel one fine night on the desolate stretch of coast road between Paekakariki and Pukerua Bay. This is the only thing that has gone wrong with the car during the eight months that I have now had it on the road.

I realised a long time afterward that there was no need for all this trouble. I could have used the VW gauge without any recalibration by including a resistor to limit the voltage supplied to it to less than 6v. Could have saved a lot of trouble.


The three pieces of glass on the Road Rat are glued straight on to the heavy hollow sections that make up the window frame, and double as a forward roll-over bar. I approached this part of the build with some trepidation. For one, it was really handy to have no glass in the window because this gave full access to the wiring, pedals and steering column without needing any Houdini acts. For another, I was a little unsure about the idea of the glass just held there by glue. I need not have worried on the second count. I made hard-board templates and had Autoglass cut the shapes out of the right sort of laminated glass. They also sold me two tubes of the right sort of adhesive, and gave me small amounts of two other chemicals and swabs needed to do the job. They also gave me a brief tutorial on how to do it right. First off I tried sticking a piece of old steel onto an old windscreen I had lying around. I used a contact area about 25mm square, and minimal preparation. Within seconds it was stuck like the proverbial. Not a chance of moving it. That took away any doubts I might have had, and I proceeded with confidence, using about twice the recommended contact area and generally being very conservative. This is great stuff. I would be happy to stick a whole car together with it. Maybe some day I will. This cost about $350 in materials, and the glass weighed 13kg.


I made the seats by gluing foam of various densities and shapes onto pieces of plywood and then wrapping the resulting squabs in pieces of brightly coloured vinyl cloth. I stapled the covers on without stitching. The idea I had here was that on occasions they will get soaked with rain, and my experience has been that when this happens, the water gets in through the stitching, so I have tried to make it so that there is no downward path for the water that will lead into the padding of the seats. I did a lot of trial and error work here to get good supportive seats, and they did work out quite well. There is about a 50mm step out for lumber support and about 100mm rise at the front for thigh support. It could have done with more in the thigh area, but I wanted also to be able to stand up on the floor ahead of the seats, to make getting in and out easier, and I was able to achieve this. At this stage the lateral support is very poor, so unless you have a substantial passenger the car is a bit slow round right hand corners. I will see to this inadequacy soon. An added complication to the seat mountings in that the firewall is designed for quick removal (release two clips and out it comes) as this provides the best access to the engine for checking oil etc.

For seat making, I found the best type of foam came from an old latex rubber mattress. This was much denser and more supportive than the various types of plastic foam I had gathered up for this purpose.

 Weigh-in time

I suspect we all secretly hoped that our cars would turn out a bit lighter than they actually did. I was pleasantly surprised to find that I was almost bang on the nail with my target of 600kg. It weighed in at 585kg ready to drive away. This is a 33/67 split front/rear. Now I look back and wonder where I could have saved a bit of weight. The frame is a bit heavy as I used RHS rather than the thinner walled ERW square tube. This could have saved 40kg or so. The windscreen and frame could have been done without. This would have saved another 30 kg or so. The next saving would have to be in fabricating wishbone suspension and using lighter hubs, brakes, wheels etc. and now we are really talking about quite a different sort of car. But with more of a weight conscious focus, a car like this one could be built under 450 kg without too much trouble or additional expense.

 The on-the-road experience

First time out, the term "all-over-the-road" certainly sprang to mind. The initial tech committee drive testers were very brave men. However, after a series of small adjustments to wheel alignment, caster angles, tyre pressures, etc. it has become a much more relaxing drive. A fair bit of this improvement seems to have come from the tyres wearing in. However, it certainly remains a lively car with a very responsive chassis.

With a 2.1m wheelbase and all the weight concentrated at the centre of the vehicle, the polar moment of inertia is very low. This means that there is very little effort required to actually get the vehicle to start turning, and then to straighten up again. Try twisting a large suitcase round by itís handle compared to say a bucket of sand of the same weight, or carrying a long ladder over your shoulder round corners compared to a sack of flour. So a long car with weight at its ends needs time to transition into and out of a corner. With a car like the Road Rat much more of the available grip on the road can be put into resisting the centrifugal forces during cornering and is not wasted on just twisting the car around.

The tyres do seem very critical. I would like to experiment with this a bit further. The car currently has wider rims and tyres on the rear than on the front. All tyres are the same profile (60 series). I suspect that a more compliant steering response can be achieved with a car like this (with more weight on the rear wheels than front wheels) by having the rear tyres with a lower profile than the front. This would need a larger rim diameter to maintain overall transmission ratios and ride height.

Ride is very smooth. The torsion bars are quite soft and the suspension has a lot of travel available. Passengers are usually surprised at the comfort of the experience. The car makes a joke of most "speed bumps" which can generally be taken at 40km/hr with no drama at all. Again, the low centralised mass and sitting position means that there is no long period dominant pitching frequency for the speed bumps to excite. The ĎRat just drives quickly over the little hill.

The car has high sides, with frame members at shoulder level as you sit in the car, and the deep three piece windscreen adds to the cosseted feeling once you have clambered in. Wind and rain are not too much of a problem. At highway speeds the rain tends to blow over the top of your head, but a certain amount of it does drop into your lap from the top of the windscreen. Noise is there too. Once the old air cooled VW engine gets up over 3000 rpm it gets pretty rowdy. And, itís only about 300mm and one sheet of aluminium away from your kidneys.

Off road it goes well too. Getting first and second at the last two club grass gymkhana events was a lot of fun. Trying it at speed on a really rough track I found the main trouble was keeping my feet on the pedals. They keep bouncing up in the air. On soft sand it gets further than most, but it isnít quite a beach buggy. The first time I got a bit stuck in sand the car gave me another one of its unique experiences. I put it in reverse, revved it up and dropped the clutch. The whole world suddenly went grey, as these two rooster tail fans of sand came flying up both sides of the car from the rear wheels. Sand everywhere.

 Exhaust system

There was a problem to overcome here in that I was keen to get all the pipes into one tailpipe, but there wasnít a clear path to get the left ones over to the right side of the car (or vice versa) without compromising ground clearance. I was keen to keep a relatively smooth and invulnerable underbelly. So I read up a bit about exhaust systems in the library and learned about the "independence" approach. This effectively uses four straight pipes from each exhaust port into a relatively large volume receiver. This apparently gives the advantages of four "tuned" pipes, but without the "peakiness". To be effective the volume of the receiving chamber has to be at least 12 (I think) times the displacement of one cylinder of the engine. So I built a six litre chamber across the rear of the car, and fed the four pipes into it, then followed that up with a small Coby muffler. The four straight pipes are about one meter long, dictated by the length of the engine/transmission, which is probably a bit too long for a serious performance improvement, but the system seems to work well. It certainly is quiet and it sounds OK. I have seen similar systems on racing side-car rigs, but with primary pipes about half the length that I used.

Stuff still to do

I got the car on the road with just the bare necessities to make it comply with the legal requirements, and still give me a bit of room to move with further developments. Since then I have smoothed off the rear body work. And I have some more "tidying" like this to do. I also have a newer engine to fit, and a priority still remains to give some attention to the gearshift linkage which is still diabolical. Then there are the seats to improve, inner guards to make for the rear (to keep out the water, dust and grass etc. and reduce the "under the lawnmower" look inside the rear compartment), and possibly some sort of emergency roof.

 Looking back

Iím happy with the way the car has turned out. The experience has been very rewarding and the car has turned out to be entertaining, reliable, practical and versatile. However, I would certainly recommend to anyone building something of a similar concept to base it on parts from something a lot newer. Just about any FWD car from the last decade or two is a good point to start. And most have at least twice the power of the old VW.

The thing I would most like to say is to recommend to others who are scratch building to stick to it and get those cars onto the road. The scrutineering system that the club has put in place works very well indeed (thanks very much to the continuing efforts of the dedicated few) and the process for getting VINed certified, inspected, WOFed, registered and let loose on the road all works well also. SoÖ donít muck about, get out there.

Brian Worboys 5/2000

.Click here to go back to the main Road Rat page