1921 Ner-a-Car

When I was a student there were four old vehicles belonging to the Student Union. To a large extent my interest in old engineering stems from my involvement in them. The one I spent most time on, and still do, is the 1916 Dennis Fire engine "Jezebel". The other vehicles are a 1926 Morris Truck "Clem"  with which I have a peripheral involvement and the 1902 James and Brown car "Boanerges". team Bo also have a 1926 Ner-a-Car which is called "Derrick" (Bo & Derrick, you see, if you are old enough).

I did once try to ride Derrick to Brighton following the Brighton Run, but he/she/it broke down half way and was  loaded onto the back of Clem under a tarp. Amusingly a chap by the name of Andy S  (of which more anon) managed to do an engine rebuild under the tarp on the back of the moving truck, and then unloaded the Ner-a-car and continued the journey. 

The IC Ner-a-Car was my first exposure to the marque. They are a most odd device, designed by an American by the name of Carl Neracher to answer a question that I suspect that no-one was really asking. They are basically a motorcycle built like a 1920s car, with a ladder-frame chassis and a fore-and-aft engine. I will not describe them in more detail here, the Wikipedia Page is a good start. 

I had, since I met Derrick, had a vague urge to own one. But an urge so vague that I never actually bothered to look for one until about 20 years later when the Andy mentioned above, who had in the interim collected a number of boxes of Ner-a-Cars equivalent to about 7 complete machines, but not actually including any complete machines decided that one of them, a 1921 Series A, was too far gone for him, and that I should buy it. 

After paying an undisclosed sum, I was soon the recipient of a large number of parts, adding up to almost a complete bike, missing only several important parts, and with duplicates of others. 

Time passes.... In fact quite a lot of time. I took delivery of the parts collection in August 2008. Numerous other projects came and went. In March 2014 an email out of the blue arrived asking if I had spares of certain parts (someone who had been in contact with Andy S). Suddenly I remembered the unfinished project, and looking though the parts decided that it wasn't as far gone as I had remembered. 

That is not to say that it wasn't fairly far-gone. Andy, when persuading me that I wanted it, did mention that I was the only person he could think of who would be prepared to de-rivet the chassis to fix the corrosion problems. It turns out that he was right. It took a couple of hours of drilling, grinding and cold-chisel work (mainly cold chisel work) to extract the 100+ rivets, and then I was left with a sort of rusty Airfix kit. 

What this picture doesn't really show is the nasty state of the rear-ends of the chassis rails, the dreadful state of the rear chassis stiffeners (aluminium) and the rust on the curved front sections of the chassis. So, here are some close-ups:

The stiffener is bad to the point of absence. 

I was faced with quite a large de-rusting project, and being rather lazy I searched the internet for an easy way to do it. The most appealing of the methods suggested was soaking the parts for several weeks in Molasses. I have no idea how this is meant to work, and I was half-expecting that I was the victim of an elaborate internet joke. However it seemed worth a try, as at least treacle is easier to dispose of than phosphoric acid. 

I build an MDF coffin for the parts, and lined it with polythene. The tank was then loaded with parts and topped up with a 10:1 solution of Lyles Black Treacle. I was quite glad for the Self-Checkout lane in the supermarket when I was buying 12 tins of treacle and nothing else:

A couple of weeks later I pulled it all out, and it had worked, to an extent. In fact there was shiny silver metal in some places, and the rust that there was seemed reduced and firmer. I wanted to get on with the chassis parts, so those were removed. The solution was decanted into a big plastic drum, and I lobbed the other parts in for a further few weeks of soaking. 

I decided to start on the chassis reinforcement first. Originally these were aluminium castings, and I did consider having castings made, but as a pattern is very nearly as much effort as the part, I decided to CNC machine parts to the same design as the originals. 

I have to say I am pretty happy with how they came out. Once painted black with the rest of the chassis they should look right. 

The chassis was made of 12 gauge steel. Which is rather hard to get hold of nowadays. (2.5mm and 3mm are what you now get). Luckily my dad had some, so I was able to use an authentic thickness. 

Not that very much of the chassis is the thickness it was made at any more. 

I made a wooden former for the new chassis ends (handily it can be flipped to make the mirror part). 

Much bashing later I had a new chassis tail-end. The CNC mill made short work of adding the right slots, and I TIG welded it in place. To ensure alignment I used a couple of extant mounting holes as references. By bolting down the rail to a piece of wood and drilling holes in the new part to match the stiffener I was able to jig things quite well for welding. 

This is rather satisfying before-and-after:

At the front I did much the same process, cut out the rusty parts:

Make a non-rusty copy:

And weld it in:

The other side, for some reason, was in rather better condition, and that just needed some of the bigger corrosion pits filling up with TIG weld and grinding flat. 

.... Time passes. (actually, only a few days)

Further up the blog I mentioned throwing some parts back in the molasses. I don't know if the solution matures or something, but when I went magnet-fishing for the parts still in there the results were a lot more impressive. (I think they got an extra 2 weeks). 

This picture probably needs a bit of explanation. The part on the extreme right was rusty, but not pitted. That is basically silver metal all over.  Further to the left is the steering column tube, dried and rubbed with Scotchbrite, which has largely smeared brand-new rust on the surface. Note how the remaining original paint is untouched. Then we have a few still-wet parts after a scrub. These are shiny on the high spots and dark in the pits. If you had seen how they went in, this is a real result. 

The second part from the left shows what happens to parts as they dry in the sun, you get an instant superficial rust. Have some primer ready if you try this. 

The next problem was that one of the engine mounts was broken. This picture shows the main casting, with the broken-off part on top. 

Luckily whatever random 1920s alloy they chose is weldable. (unlike, for example, Dennis N-type crankcases, that even competent welders can't weld). This isn't great welding, I am not a great welder.

But my new toy, a electric file (or very narrow belt sander) hides a multitude of sins. 

Another new toy, a variable-density automatic welding mask with a big window has also proven to be a huge help, I can see so much better with more light. Try searching eBay for IQ1700. I am rather pleased with it. 

My repairs to the rusty sections had removed several important rivet holes, so I started to test-assemble the chassis with screws and nuts. As might be expected, the majority of the missing holes were on the outside meaning that the trivial solution of drilling through a mating hole into the missing hole was not an option. I solved this problem by using a world-changing technology called "paper". 

By inserting a piece of paper in the gap and rubbing through the holes, or in some cases simply cutting a piece of paper to approximate shape and pushing screws through the paper and several reference holes, plus the holes that needed to mate to missing holes the hole-de-removal took rather less time and effort than I expected. I trial-assembled the entire chassis with screws:

I ran out of red primer, but had some forgotten grey in a corner....

Some of the rivets are not accessible with the chassis assembled, so it will need to come apart again, but this looks like progress to me :-)

I am still trying to decide how to do the rivetting. In many cases the normal manual way won't work, there is no access at the back for a rivet-pop. I experimented with a pneumatic hammer, but that can't actually do the job with rivets of this size in steel. (This was a £12 pneumatic Screwfix hammer, quite a good tool, and I thrashed one making mullion windows out of sandstone in a previous project, and it lasted years). I have a 12-tonne hand-held cable crimping press on the way. If that won't squeeze the rivets then I probably need to make a set of round-the-corner rivet-pops, which will be tedious. 

Nobody ever dismantles a working motorcycle

This is a truism echoed by anyone who has ever restored an old vehicle from parts. At some point during your restoration you will find out why your  barn-find was parked in the barn in the first place. 

I am not saying that I have found the only thing wrong with this Ner-a-Car, but I found one of them. The front swinging arm was bent. By really quite a lot. I guess that the wheel would have sat at a 20 degree angle: Ideal for right-turns, less so for left turns. 

They may have tried to fix it, and given up. Attempts to operate the the pin-spanner holes in the front hub with a hammer and pointed stick had not only damaged the holes:

But had also peined the threads into the hub, making them really hard to undo. 

I made a special tool:

And then used some extreme measures to hold the hub and keep the drive pins from camming out. 

This allowed me to dismantle the hub, which is remarkably simple in design compared to a Bimota Tesi or a Vyrus. It did not need to bother with handling braking forces, nor did it have any real suspension as we now understand it (springs but no damping). In fact it didn't even have to be any good at all. Which was lucky.  

The bearing is metric, 50 bore, 90 OD and 25mm wide double-row ball bearing, made in Switzerland. 

That's no longer a standard size, you can get 50/90/23 but not 50/90/25. I think that in practice you could use shims to fit a modern bearing, or possibly there is enough adjustment in the threaded keeper plates. I won't go into more detail here, there is plenty of info on this brilliant website .
I was concerned about the bearing, and availablity as the bearing felt horrible. In fact it would only turn one way, and was notchy the other way. 

Luckily after a good clean and re-grease it feels OK. Not great, but OK. 

Once I had it all apart I could try to fix the bend. 

I just managed managed to get it hot enough to straighten it by hand using both blowtorches and an assembly of vermiculite tiles round the area to retain the heat while holding the pivot area in the vice. 

While I had it all apart I decided to fix the pin-spanner holes for future use, so I filled them with TIG weld:

The re-drilled the holes:

The distorted threads were also a problem. I managed to re-chase the threads on my CNC-converted lathe using an interesting procedure. 

1) Set up the parts on the chuck. 

2) with no tool in the toolpost run the threading macro until it is on the spring-cuts (the last ones at full depth). 

3) Turn the spindle-override to zero. 

4) Push a threading tool deep into the existing thread. Then stop the cycle. 

5) clamp the tool in the toolholder and run the threading macro again. 

It seemed to work, the rings now screw in, and out, and lock with their screws. It is possible that I might end up just making new rings in the future, but originality counts for something (and I had no bar-stock big enough). 

Next I will be trying some wheelbuilding, and also making a rivet squeezer. I do hope it works.