Gas, tires, oil.

I wanted to know what my ‘68 Tempest had for a rear axle ratio, I’m thinking about using it for commuting and am wondering what I could do to improve mileage. First I jacked up one side of the rear axle (if you have positraction you’ll need both wheels off the ground), then put a chalk mark on the drive shaft and the differential.

Then I put a mark on the tire where it lined up with a crack in pavement.

Next you just rotate the tire, and count how many times the driveshaft rotates and the tire rotates until both marks line up similtaneously. It takes a lot of turns, I kept tally on the sidewalk with a piece of chalk.

It took 32 turns of the driveshaft and 25 turns of the wheel before the marks lined up again. 32/25=1.28, but since only one wheel was turning I have to multiply by 2 to get 2.56 for the actual ratio. I guess I’ll have to look elsewhere for gas mileage improvements.

Over the years people have used various schemes to measure the aerodynamics and parasitic drag of a vehicle using what is called a “coast down” technique. What you do is take a car up to speed, put it in neutral and let it drift. You can then use a stop watch to measure the time it takes to decelerate a certain amount (say, from 65 mph to 60 mph). Given the weight of the vehicle, you can calculate force opposing the vehicle at this speed and the horsepower needed to maintain that speed. If the measurement is done at several speeds, you can separate the rolling resistance (tires, wheel bearings, brakes dragging, etc), which is proportional to speed, from the air resistance, which is proportional to the square of velocity. Obviously, you need to find a section of road that is flat and level, and you should ideally avoid windy days, though repeating the test in opposite directions would help.
What I would like to develope is a method to record continuous deceleration data. This would allow you to fit the data and obtain much more accurate results. These sorts of results could be very interesting, for instance, you could determine the effect of various modifications on performance. Is it better to leave the tailgate up or down on a pickup, is the drag from a rooftop luggage carrier enough to make it worth removing when you are not using it, what is the change in rolling resistance with tire pressure? Now, I can think of several ways to obtain this data, first, you could use an inertia detector to directly record the deceleration. I believe there are commercial devices designed to measure acceleration performance in just this manner and I wonder if the hard drive protection inertia detector in mac laptops could be used. Another method would be to take data from the computer of a modern car through a laptop interface. One of the data channels is speed, though you would be limited by the accuracy of the speedometer. Having neither a mac nor a modern car I decided to use my GPS (a garmin 12map) and pc laptop. While gps reciever have a well known error in position, the speed seems to be very accurate. In addition, they also indicate elevation, which may be useful in finding a flat piece of road for the test.
Most GPS recievers can ouput real time data through a serial cable. Mine can do so with either a proprietary Garmin protocol (which is documented) or a standard called NMEA, which is plain text. I searched around for couple days for an interface program, but was unable to find exactly what I wanted. Plenty of programs record position, but not speed. So, I decided to just log the entire output, then pull out the data that I wanted. I used Hypertermial (comes with Windows) to connect to the gps. You need to specify Com port (com 1 for me), 8 bits and no parity and 4800 baud. If successful you will see a bunch of text scroll by on your screen. You can either cut and paste from the screen (set the buffer way up) or log the session to a file. Here is some example output:
$GPBOD,,T,,M,,*47
$GPRTE,0,1,c,*36
$GPRMC,024433,A,3958.908,N,13148.360,W,025.8,358.9,261106,015.1,E*6B
$GPRMB,A,,,,,,,,,,,,V*71
$GPGGA,024434,3658.910,N,12148.360,W,1,09,1.4,79.9,M,-28.9,M,,*4B
$GPGSA,A,3,,06,07,10,16,18,21,,26,29,30,,2.4,1.4,1.9*38
$GPGSV,3,1,11,03,00,315,00,06,63,140,49,07,67,305,47,10,07,043,38*7D
$GPGSV,3,2,11,16,26,298,43,18,58,214,38,21,63,323,46,22,20,219,00*79
$GPGSV,3,3,11,26,33,089,42,29,30,072,42,30,08,185,34,,,,*41
$PGRME,3.7,M,4.6,M,5.9,M*24
$GPGLL,3658.914,N,12148.360,W,024434,A*32
$PGRMZ,262,f,3*1D
$PGRMM,WGS 84*06
$GPBOD,,T,,M,,*47
$GPRTE,0,1,c,*36
$GPRMC,024435,A,3958.922,N,13148.360,W,024.6,358.5,261106,015.1,E*66
$GPRMB,A,,,,,,,,,,,,V*71
$GPGGA,024436,3958.923,N,13148.361,W,1,09,1.4,80.1,M,-28.9,M,,*46
$GPGSA,A,3,,06,07,10,16,18,21,,26,29,30,,2.4,1.5,1.9*39
$GPGSV,3,1,11,03,00,315,00,06,63,140,49,07,67,305,47,10,07,043,37*72
$GPGSV,3,2,11,16,26,298,43,18,58,214,38,21,63,323,46,22,20,219,00*79
$GPGSV,3,3,11,26,33,089,42,29,30,072,42,30,08,185,34,,,,*41
$PGRME,3.9,M,5.0,M,6.3,M*24
$GPGLL,3958.927,N,13148.361,W,024436,A*31
$PGRMZ,262,f,3*1D
$PGRMM,WGS 84*06
The data we want is in $GPRMC (GPRMC = global positioning recommended minium navigation information). Next time, I’ll show a Perl script that I wrote to pluck out the time and speed, and an example calculation. But here is a little preview just to show that this might work (i.e. data is recorded at a rate sufficient to fit):

I recorded this with my suburban on the road outside my house (you can see effect of couple small hills near the end coast down section and a car was coming up behind me so I didn’t come to a complete stop), but even at 40 mph you can see the effect of air resistance (the speed falls much faster per second at 40 than at 10 mph), but of course the suburban has the aerodynamics of a brick.

Next up on the trike was the front brakes. I had new brake lines to replace the old corroded ones, but the open end wrench just slipped on the nut. Partially rounded off fittings like this are a common problem, I wasn’t re-using the line, so I didn’t mess around and just cut it off.

Now I was able use a hammer to drive a six point snap-on socket over the abused nut and then turn it off with a 1/2″ drive 18″ combo ratchet/breaker bar.

Yes, I have 1/2″ drive sockets down to 3/8″ (and 1/4″ drive sockets up to 9/16″), and, as you can see, they are very useful. What if the 6 point socket slipped? I would have driven a 1/2″ nut over the fitting and then welded it on. This not only give you a fresh, larger hex to grab, but the heat helps loosen it up (but is much preferred to trying to heat it with a torch and possibly damaging the brake seals).

Just about everyone who works on cars and bikes hates electrical stuff, but you have to do it. I have grown to hate the common insulated crimp-on connectors. Firstly, they let water in and become corroded. Next, they are really just flattened or squished by the crimping tool, this allows the wire to wiggle side to side and eventually come loose. And lastly, they just look terrible. So, here’s what I do (I be glad to hear of other suggestions). I use non-insulated connectors (below, bottom) or if I can’t find them, I will cut the insulation off of the insulated type (top).

Not all of the common crimping pliers are made to do the non-insulated ends, mine have it on the handle side of the joint. Make sure you put the split side of the connector toward the half-round mandrel of the crimp tool.

Now, that wire is not going to pull out.

Next I use heat-shrink tubing to insulate and provide strain-relief. If you can find it, they make heat shrink tubing that is coated internally with hot-melt glue, this will help keep moisture out. I then give it a coating of liquid electrical tape (may be called brush-on electrial tape). I beleive it’s just plastisol (the stuff they use to dip-coat plier handles).

If your connection will show (and you care), the liquid electrical tape is available in different colors (nice for battery terminals if your cable is black, you can use it to mark the positive red).

The old style harley manifolds are a pain to install, an extra pair of hands can be a big help. Here’s one way to do it alone. There is half a groove on the manifold and the other half on the head, so the o-rings won’t stay on either part. So, first roll the o-rings up on manifold past where they should be.

Now you can roll one side down in place without the other falling off.

Install the clamp on that side to hold everything in place.

Now you roll the other side into the groove.

Of course the better way is to not use the stock o-rings at all. You can buy spacers that can be glued to the head and the manifold in the o-ring groove. This allows you to use the more modern band type seals and band clamps.

I needed some rubber washers to mount the gas tank on the trike, but the hardware store didn’t have the right size. I decided to make some myself. First I found an old inner tube that I had stashed behind the workbench, just for this sort of thing. Then I got an arch punch that I had found at a yardsale. I quickly punched out the ODs.

The ID needs to fit a 5/26″ bolt, but I don’t have that size punch. However, when at a recycling center I noticed some empty cartridges in their scrap brass bin. I bought a selection at scap price (and got the bucket for free).

5/16″ is .3125″, so 8 mm at .315″ is a nice fit. This is the cartridge head of an 8mm mauser (or 8x57mm) that I found. If you want you can sharpen the end, this also allows you to vary the size slightly, you can sharpen the inside (with a small round file or chamfer tool) for more clearance or the outside for less. So what cartridges should you look for? Well, obviously 6mm, 7mm, 8mm, 9mm and 10 mm will work for the corresponding metric sizes. An inside sharpened 17 caliber should work for a #8 screw and outside sharpened it should work for #10. The new 204 ruger might work for 5mm or #12. 25 caliber (25-06, 257 roberts, etc) or 6.5 mm for 1/4″ bolt. 8mm for 5/16″. 375 caliber or 9.3 mm for 3/8″, 44 or 45 caliber for 7/16″. And of course, 50 caliber for 1/2″. Good luck finding a 600 Nitro Express for 5/8″. Also, some cheap military ammo uses steel cases, obviously these will be sturdier.

I didn’t bother to sharpen the end and it worked fine.

Here are the rubber washers in place on the top tank mount, trapped between stainless washers. I have previously shown the bottom mount.

And here are the fat bobs finally mounted. The dash and filler are just sitting on top, they’re next.

The take away message here is you can’t just smear grease on a bearing. You have to force grease between the rollers. They make a tool so you can use a grease gun, but doing it by hand is much more fun (and satisfying). First, get a big gob of bearing grease on the palm of one hand.

Then, in little bites, scrape the grease off your palm with the outer edge of the bearing’s roller cage. Here I’m doing the front wheel bearings for the dana 44 I’m putting in the suburban.

You know you’ve got it when you can see grease come out the top of the bearing like above. Then rotate the bearing till you get the whole thing.

I put a new fan belt the suburban today, the old one wasn’t broken, so I put it in the box the new one came in and tucked it inside the front fender.

There’s a surprising amount of space inside a fender, I’ve tucked old radiator hoses, tuneup parts, belts, whatever fits. And they’ve come in handy. One time, I hit a really big pothole up the the national forest, hidden by snow, that cause the fan to slice the upper radiator hose. Luckily, I had an old one in the fender, other wise it would have been a long cold walk (20 miles of deep snow). And the box? It has the part # on it, so when I need a new one I can be sure they give me the right one.

One of the few things I got done over the weekend was changing the oil in my sportster. I like to check the old filter for metal particles as a diagnostic tool. You need to cut it open because the oil flows from the outside in. First, pop a hole in the side, near the top, I used a screwdriver.

Then I use a nibbler to cut all the way around.

(more…)

Most american screws (non-metric) under 1/4″ are named by their number and thread pitch (thread per inch). So a 10-32 and a 10-24 are the same diameter (#10), one is fine thread and one is coarse. To calculate the diameter you multiply the number by .013″ and add .060″. A #10 is therefore .060 + (10*.013) or .190″ (about 3/16″). I plugged the formula into a spreadsheet and calculated some common sizes.

Sure, you can look it up online or in Machinery’s Handbook, but if you know the formula (and even a lot of machinists don’t) you can calculate the screw diameter anytime.