Imperial Journey

In the last article, Hoodie 2, we finished working on the hood. Which meant it was time to move on to the rest of the car.

Paint serves many functions. It is the most visible part of a car, driving your perception of the car. It is a protector, keeping the underlying metal from rusting. And it is an opaque barrier, hiding whatever is beneath – whether it is solid metal or Bondo… This last part is critical with old cars.

I’ve already covered several discoveries hidden by paint and trim. And strongly suspect that there are more hidden surprises lurking under that sinister black paint. There are several bad areas in the paint – it looks like something was left on the hood which ate into the paint:

Several spots on the trunk lid are down to primer:

There are areas where the paint is badly cracked:

And most interesting are the numerous speckles on both sides of the car – with the exception of the front passenger door, which we know was previously repaired. These speckles go completely through the paint, there is grey metal at the bottom of the holes that looks like lead, and the holes haven’t rusted. My theory is that the car was shot with a shotgun – repeatedly. I would really like to know more about the history of this car!

Between the known issues and the unknown issues the only way to prepare the Imperial for paint is to completely strip the entire car down to bare steel, repair as needed, and build up a foundation for new paint. I have to become a stripper – a paint stripper!

But how?

An orbital sander is much, much too slow. A flap disk takes paint off quickly but damages the metal. Chemical paint strippers are incredibly messy and don’t work well – EPA regulations have made paint stripper less toxic and less effective than it used to be. What is left?

Research turned up something called strip disks. These are basically a sponge like material with abrasives embedded in the plastic and are designed to go on an angle grinder. The theory is that they are abrasive enough to remove paint, the open mesh keeps the abrasive from getting clogged with paint (an issue when using sandpaper), and they don’t damage the underlying steel.

Worth a try, so a box of BHA Easy Strip Discs was ordered. Many choices are available; this one was chosen based on a combination of good reviews and a mid-range price. My theory is that the cheapest choice is cheap for a reason, the expensive choices may be either good or simply over priced, and that medium price ranges are often the best balance of quality and cost.

With a strip disk mounted on the angle grinder, an N95 face mask strapped firmly in place, and safety goggles perched on my nose I cautiously approached the Imperial, ramped the grinder up to 11,000 rpm and applied it to a test area.

The strip disk worked great! It quickly and completely removed the paint and left the underlying metal undamaged. In fact you could still see the various imperfections in the surface of the steel. This will work!

Well, “quickly” is relative. We are talking about removing paint at the rate of square inches per minute. A single door takes about half a day, and a fender takes half a day to a day. Reaching the center of the roof and trunk lid is a major stretch – and the grinder gets heavy very quickly when waving it around at full extension of your arms. Bracing yourself in position so that you can stretch that far uses leg muscles that aren’t used for any other purpose – as they remind you for the next three days… Stripping paint off of a large care is literally a pain!

The bad news is that it took over a week of incredibly dirty days and two boxes of strip disks to complete this miserable task. I ended each day looking like a racoon once I took off the face mask and safety goggles. I left my clothes in the basement to avoid tracking dirt through the house and had to run them through the washing machine twice to completely remove the paint dust from both the clothes and the washer. The water going down the shower drain looked like something out of an Alfred Hitchcock movie – in black and white, of course!

The good news is that the paint was completely removed, giving me a solid foundation for a good paint job.

The results?

The strip disk turns the paint into paint dust creating a massive mess. How massive? As previously mentioned this car has about 64 acres of surface area. It requires about a gallon of primer and 1-1/2 gallons of paint to cover it, or about 2-1/2 gallons total. Automotive paint is roughly 80% solvent and 20% solids. This means that there is about 1/2 gallon of solid paint on the car which is turned into dust. Plus any Bondo that was hidden under the paint – and Bondo is applied much thicker than paint. Of course dust takes up more space than the equivalent solid, so we are looking at creating 1-2 gallons of paint dust – which is blown off the car at high speed by the angle grinder. This dust ends up EVERYWHERE! On the car, on the floor, in the back yard where it has been sucked out by the vent fan, on the protective curtains, on the shop tools, and everywhere in the shop.

Not to mention in the shop vac. The trusty shop vac was used almost continuously to try to manage the paint dust. The car and floor were vacuumed repeatedly. I emptied the shop vac and cleaned the shop vac filter at least 4-5 times during this project.

With all of the paint removed from the Imperial and the car, floor, table, shop equipment and other miscellaneous surfaces vacuumed it was possible to take stock of the actual condition of the body.

Uh Oh – what is that at the bottom of the quarter panel previously hidden under a layer of bondo?

Surprise! One of the worst ways to “repair” rust – just pop rivet a piece of sheet metal over the top of the rusted area, cover with Bondo, and paint to match. Plus other rust holes that had just been Bondo’d over.

Actually, this was almost good news. I had seen evidence of this patch panel on the bottom of the quarter panel and was afraid it covered a much larger area. Stripping off the paint showed more rusted areas on the other quarter panel, although without the nasty patch.

This whole area will be cut out and new patches welded in place. Only a skim coat of body filler will be needed, unlike the 1/4″+ applied over the riveted patch. It will take time to patch these areas properly – but it will be done properly this time!

Other details were also revealed. Several small dents turned up that need to be fixed as well as a couple of larger dents. The shotgun blasts will need work. I’m not nearly as close to being ready for paint as I had hoped to be.

But, overall, the car is solid and in good shape. There were fewer surprises than I really expected. The rusted areas hidden under the paint are in areas that are almost guaranteed to rust on older cars. There is surprisingly little rust in the front fenders and the rocker panels are solid. Even though I’m complaining about what was hiding under the paint, I really can’t complain.

Just to add to the entertainment, I won’t be able to work on the car for several weeks. It can’t be left as bare steel as it will start to rust again. Since there is paint dust everywhere I had to vacuum everything, remove all of the old masking, vacuum again, wash twice with solvent, mask everything again, wash with solvent, go over it with a tack rag, and spray a coat of primer. Which I did.

The epoxy primer leaves a flat back finish that is popular today, especially on hot rods. It looks positively evil on this car, and I mean that in a good way. I’m tempted to black out the chrome trim, put it back together, and drive it like this! I’m not going to, but it is tempting…

When I get back to work on the car the plan of attack is rust repair followed by dent repair. I have to figure out how to deal with the shotgun blasts – hopefully high build primer will take care of this. The whole car will be sprayed with high build primer and the surface perfected. Once the surface is perfect a final coat of epoxy primer will be applied and then the car will be ready for the paint shop.

Next: Wrangling Rust

With the bottom of the hood done in Hoodie I flipped the hood over and tackled the top. The top of the hood has bothered me since I got the car – something had been laid on top of the hood that attacked the paint, there were areas where the paint was badly cracked, and multiple large stone chips. Basically it looked horrible.

The hood needed to be completely stripped of paint, skim coated with body filler, and primed. The DA sander I used to strip paint off the door in Door Four Debacle worked but was very slow. A flap disk on an angle grinder quickly stripped all the paint off the hood – but at a cost. Even though I was careful the flap disk noticeably gouged the surface, leaving marks that would clearly telegraph through the finished paint. Fortunately I was already planning to skim coat the hood, which will hide the gouges. The flap disk is clearly too aggressive for paint removal. Something else is needed for the rest of the car – hopefully something faster than a DA sander…

With the paint gone a thin coat of body filler was applied and worked with the body board to produce a completely flat surface. Here you can see the high areas have been sanded down and low areas remain.

Working large areas of filler with a body board is a huge amount of work. After a couple of hours of back breaking (and shoulder breaking and biceps breaking and wrist breaking) work to get partially done I called it quits for the day and headed inside to order power tools.

A jitterbug, more properly a straight line air sander, is the tool of choice in body shops. A jitterbug is fast – too fast if you don’t know what you are doing. The general recommendation is to use a manual body board until you learn how to work filler. I decided that I had learned enough with the manual board and was ready for the pro tool!

The jitterbug loaded with 36 grit sandpaper made short work of rough shaping the body filler. The bare metal represents the high spots and the filler has filled in the low spots. The hood was already fairly flat, so the filler is less than 1/16″ thick. After finishing the shaping with 36 grit, 80 grit sandpaper was loaded onto the jitterbug to get rid of the deep scratches left by the extra coarse 36 grit sandpaper.

With the rough shaping done the next step was to perfect the surface. Two coats of epoxy primer were sprayed onto the hood to seal it and provide a foundation for high build primer. Two or three coats of high build primer were applied, followed by a light guide coat. The DA sander was loaded with 180 grit sandpaper and sanded until all of the guide coat was gone.

The main purpose of the high build primer is to fill in the sanding scratches left by the 80 grit sandpaper in the previous step. After sanding with 180 grit another guide coat is sprayed and a final pass with 320 grit sandpaper to remove the 180 grit scratches finishes sanding.

You might notice that the DA sander has two hoses attached to it – an air hose and a mystery hose. The large hose is attached to a shop vac and pulls sanding dust through holes in the sanding disk. This new sander greatly reduces the amount of sanding dust released onto the air and should minimize cleanup. We will see how it works out – initial results are encouraging.

Two coats of epoxy primer seal the high build primer and provide a surface ready for the actual paint.

There is a little bid of banding in the epoxy primer because I sprayed it with a small finish gun that makes it difficult to smoothly cover large surfaces. In the future I need to use a full size spray gun for these larger jobs. The small gun is great for smaller areas and provides better fine control – you just need to use the right tool for the job at hand.

The hood came out quite nice. All that is left is the other 90% of the car… Which we tackle in I’m a Stripper…

After finishing the last door in Door Four Debacle it was time to move on to the next rust damage I had been avoiding – a hole in a rib in the hood inner frame.

Wrestling the hood off of the car is normally a two person job, but only one of me was available. No problem! Drag out the engine hoist, hook it up to take most of the weight, and I can handle the rest. Demonstrating the wisdom that comes with age (or perhaps expensive and/or painful mistakes…) I quickly realized that you have to have people on both sides of the hood to control it while it is flying through the air. Desperate times call for desperate measures, so I asked She Who Must Be Obeyed for help. With two people guiding the hood it safely migrated off the car and onto the workbench.

Unfortunately I didn’t get a before picture, but the hole was at the end of one of the pressed ribs in the hood frame, next to the hood hinge. This is a complex area: multiple curves at different angles and all edges have a large radius. A single face of the damaged area is joined to up to six other faces, and all of them have to match exactly. There is a reason I’ve been stalling on this job!

My initial plan was to make a single patch panel to replace the entire area I had to cut out. This is elegant but very difficult to do… One of the key lessons learned on other patches is that you can create a single large, complex part out of multiple smaller and simpler parts. Further, you can start with one small part, weld it into place, and then fit the next part to it – bend, fit, mark, trim, precision sand the edge to fit, and then weld this new small piece into place. Continue like this across the entire part and you can build complex shapes that are quite precise.

One of the advantages of working with metal is that you can add (weld) and subtract (cut, grind, or sand) material freely until you are happy with the result. With patience and a willingness to re-do work you can achieve surprising results. After grinding welds smooth it is difficult to tell where repairs have been made. A bit of body filler, some high build primer, and a coat of paint and the result looks perfect! For certain values of perfect…

In previous repairs I was able to bend the patches by either using my (small) bending brake or simply over the edge of my steel fabrication table. This produced a small radius curve at the bend, but I was able to fit the patches to the surrounding structure. This approach just wouldn’t work on the hood.

I needed something to match the radius of the curves. After mulling this over for a while I realized that I had an assortment of bolts of various sizes. OK, this could work.

With this approach in mind I cut out the damaged area and started to work. The first question was what is the radius of the existing edges? I grabbed a handful of bolts, cut out small test pieces of sheet metal, and bent the test pieces around the bolt. Holding the bolt and sheet metal panel for bending was an issue. The brute force approach was to simply clamp the bolt and metal part in a vise. This would be much easier with three hands… Doing it this way is slow, but acceptable for the small number of pieces needed for this repair.

A 1/2″ bolt produced a curve the nicely matched the existing hood frame. The rest of the bolts were returned to their storage boxes to make sure I didn’t get confused when making parts. Of course I haven’t actually done anything like that in the past!

With the problem of bending the edges to shape solved the job became relatively straightforward: Pick an edge to start with. Cut, bend, and fit a piece of metal to that edge, perhaps starting with a cardboard template. Once happy with the fitment, weld the new part into place.

Then choose the next edge to work on. Cut, bend, and fit another piece of metal to match this new edge and to the part previously installed.When happy with the fit, weld the new part into place. Repeat this process until you once again have a complete frame section with no holes.

The corners where the parts meet were basically spherical and presented special challenges. The new pieces being welded in were left somewhat long. After welding them in the corner was worked with a body hammer until it had the desired spherical shape. Cutting notches in the corner and careful hammer work produced acceptable corners. With more practice I should be able to do even better. Once the corners were worked into shape they were welded up – both the edges and the notches that can been cut into them. The corner was further shaped with careful grinding, resulting in a smooth and convincing result.

Compare this with the other side of the hood frame which received no repairs:

Not a perfect match but solid and much better than a rust hole.

With the rust repairs done I took advantage of the opportunity to strip the paint off the bottom of the hood. This was partially done as a dry run for working on the rest of the car. Stripping the paint with an orbital sander works but is tedious and slow. Not ideal for the 63 acres of paint on the rest of this car…

After removing the old paint the hood received two coats of epoxy primer and two coats of finish paint.

While the end result looks good, I should have taken a bit more time and applied body filler to the patched area before painting. If you look closely you can see some divots and scratches from the grinder that have telegraphed through the paint – exactly the sort of thing that body filler and primer/filler hide. I need to remind myself that this is on the bottom of the hood and that no one else will ever notice or care…

Next: preparing the side of the hood that shows in Hoodie 2.

The last electrical article looked at dash lighting in Electrical 12: Dash It All!. It is now time to deal with a sticky situation – specifically, power windows that stick going up and down.

Power windows are the among the worst cases for old car electrics – the motors require high amperage and are fed through long wires and small switches. Over time the window mechanisms become harder to operate, requiring even more power from the motors. At the same time the wiring degrades, reducing voltage available to the motor and making it work even harder. The result is that power windows simply don’t work well on old cars – when they work at all.

In my case one of the windows (drivers door, of course…) doesn’t work at all, two windows barely work, and the fourth window is slow and doesn’t always go all of the way up.

Part of the work on the doors includes cleaning and lubricating the window mechanisms. This certainly helps, but doesn’t go far enough to solve the problem.

Really solving the problem requires completely replacing the existing power window wiring with something that can handle the power requirements.

Relays will be used to provide power directly to the window motor – basically the same thing that was done with the headlights when they were upgraded. The existing switches still control the windows. But instead of handling the 20-30 amps that the motor draws, the switches will now trigger the relays and only have to handle less than 0.1 amps. The existing wiring and switches are easily able to handle 0.1 amps.

The window motor requires two relays – one for up and one for down – which makes the wiring somewhat more complex than the headlights.

Power for the relays will be provided by a new dedicated 12 gauge power line that is run directly from the relays to the fuse box. Power will now come from a shorter run of heavier gauge new wiring – the motor will now get full voltage and full power.

Since I’m paranoid – in other words, I have experience… – the relay upgrade will be done as a series of incremental steps with testing at each step. Thus the first step is to build a test harness to allow testing the window motors outside of the existing window wiring. This test harness will go through a series of modifications and upgrades as the build proceeds.

The initial test harness was simply two wires connected to the battery. There are two wires to the motor – I was pretty sure that the motor case is ground, and you apply +12V to one wire to go up and +12V to the other wire to go down. But it could also be reversing plus and minus between the two wires for up and down.

With the two wires connected to the to the battery, the negative wire was held against the motor case and the +12V wire was applied to one of the wires to the motor – and the window went down! +12V was applied to the other wire and the window went up. OK, motor wiring is confirmed.

Next a two way switch was added to the test harness. Technically this was a dual pole/dual throw momentary contact switch, also known as DPDT momentary contact. This allowed using the switch to control up and down on the motor. Connectors that mated with the motor were added. The test harness is long enough to reach from the battery to each door and to the workbench, making it easy to operate the windows either on the car or on the bench. This was convenient since you need to move the windows up and down to remove the glass. As previously mentioned I’m paranoid, so a fuse was also added.

Since the test harness will also be used to implement the new dedicated power wiring, a matching power connector was added. The test harness can be used to directly power and control the window motor in the car or on the workbench as well as provide power to the relays.

With motor operation verified and the test harness ready to go it was time to build the relay setup. The first step was to lay out the design on paper, double check it, and determine how to connect the relay wires. I’m using Bosch style relays which have five wires – trigger/source, ground, power in, power out normally open and power out normally closed. Since the relay should provide power when the switch is turned on the normally open (NO) output is used and the normally closed (NC) output was sealed off.

The relays I’m using came with sealed sockets and pigtails – this makes it easy to wire them up. The good news is that the pigtail wires are color coded. The bad news is that the color coding didn’t comply with standards… Notes that include pin number, color, function, and wiring instructions kept me from becoming excessively confused when making the four relay packs (one for each window).

I wired up the relays on the bench – it is much easier to work on the bench, and it gave me the chance to double check everything before installing. Note the labels on the wires – this minimizes mistakes and makes it much easier to troubleshoot in the future.

The test harness and a voltmeter allowed complete testing on the bench. No mistakes were found in the wiring. OK, OK – no mistakes were found when the relay packs were ready to install in the doors…

The window motor was originally grounded through the door skin. Since it is very little additional work to run two wires instead of the planned one wire for power, a dedicated 12 gauge ground wire was added. This ground will tie directly into the upgraded grounding system under the dash and is bolted to the power window mechanism frame which is connected directly to the motor. This also provides a convenient high quality ground point for any other potential electrics in the door.

The new power and ground wires are connected to the relays using a waterproof Metri Pack 280 connector. The relays are sealed, and all wiring connections are sealed with marine grade heat shrink tubing – the result is that there shouldn’t be any corrosion in the power window system.

The motor originally used non-sealed Packard 56 connectors, which I replaced with more of the Metri Pack 280 connectors. When I cut the old connectors off and stripped the insulation, the wires were solid black with corrosion. Sandpaper and electrical contact cleaner cleaned up the wire ends and the new sealed connectors should prevent more corrosion in the future.

The test harness was plugged in to provide temporary power to the relays and window operation tested. It worked!

With everything ready to go I spent an hour or so making a bracket to mount the relays. I then tested window operation and watched the window hit the relays as it neared the bottom position. Actually I stopped the window just before it hit the relays. I’m getting paranoid about that sort of thing…

Studying the situation more, I decided that the relays could be mounted directly to the inner door skin. The relays were mounted in this new position, the window run down, and nothing hit. OK, one more problem solved!

The window now goes up and down smoothly. It will need to be adjusted to fit the new weatherstrip I’m installing. This will probably be a nightmare; if it is I will do a post.

This process will be repeated for the other three doors.

I also need to finish connecting the new power wiring. As I do each door I’m running the new power and ground wires up to under the dash. The next to last step will be to connect all of the new power and ground wires and splice into the existing power window feed from the under hood fuse box.

The last step is to enjoy the new smoothly operating and reliable power windows!

Next electrical: Electrical 14: MeterMatch.

Next: battling rust in Door Three “Delights”.