Overview:
Even before I bought my 3D printer, I had marveled at a series of
printed automotive engines a fellow on-line had designed and
built. His screen name is ericthepoolboy, but his real name is
Eric Harrell.
He has designed and made available at no cost through Thingiverse, all
the
design files for about 6 different engines and several
transmissions. He
has also designed a Toyota driveshaft and rear axle assembly. Now
that I am about 1 1/2 years into using my 3D printer, and after seeing
the results of a friend building one of Eric's engines, I decided I
would
tackle building one. The flathead Ford engine was my
immediate choice since I had worked on so many of these (full size)
engines in my early car-owning and hot-rodding days.
This is a significant departure from the majority of
the printing I have been doing. Mostly I have been designing and
printing specific parts for my own projects and have only printed a few
small items from others' designs. This engine is the largest
project I have yet attempted, and it is all from Eric's designs.
This is the published photo from Thingiverse, the location where Eric
makes his designs available to others. It was the inspiration for
starting the project myself.
Eric's models are beautifully designed at 35% life size. He has
gone
overboard to exactly duplicate the actual engines. Most of the
major parts are attached using magnets
which position and hold the parts very well, but allow an easy
disassembly and re-assembly. There are about 120 magnets used in
the construction of this engine. A full
hardware kit is offered by Maker Rx (makerrx.com) for just under
$100. It is worth every penny! All the screws, nuts,
bearings, threaded rod, motor, speed controller, valve springs, etc.
are
included- everything you need other than the printed parts. It
would take easily that much in your
time alone to track down all the specific parts needed!
I was also planning to
build an engine stand of his design, but the hardware kit has
been sold out since before I started this project. I have built a
quick working stand,
and am now designing my own scale model of a mechanic's engine
stand.
For my model I chose a red block with some cast iron color
showing. I did this by printing the
block in a gray color, then carefully masking and painting the majority
of the surface with a paint color called Cherry Red.
I then printed the
heads and intake manifolds (upper and lower), along with all the
other small parts scheduled to be a bright, shiny silver. I chose
a red filament for the timing
gear cover and the water pump bodies. After printing them and
installing them on the painted red block, I realized that the two reds
clashed badly, and I re-printed those parts in silver. I
originally printed the exhaust headers black, but decided they needed
more bling, so I reprinted them, also in silver.
Here are a few of the more relevant Specifications of this build:
- I initially printed 257 parts, including some trial
parts, a couple reprints, and several intentional overages. 210
parts were actually
needed.
- The total printer run time was 251 hours, 44 minutes to
print these parts - about 10 1/2 days!
- It used 1819.4 grams of filament - about 4 pounds
(numbers from my slicer)
- The largest print, the engine block, took 66 hours and 22
minutes to print - almost 3 days!
- The
gcode file for the engine block contained over 8.7 million lines of
code! That's about 145,000 pages at 60 lines per page, or 145
reams of paper double sided, but I didn't print it out.
- It took me just over 3 weeks from start to
completion. That time did include a 3 day motorhome rally, so
really just under 3 weeks!
A note about magnets:
There are about 120 magnets used in this model to hold parts
together, yet allow them to be easily removed. These magnets are
incredibly small at 4mm diameter and 2mm thick. That is well
under 3/16 inch in diameter and half that thickness. Everywhere
two
magnets
are used to hold things together, one must have its north pole meeting
the
other's south pole. As the magnets are not marked I had to come
up with a technique of easily and reliably installing every magnet
correctly. I initially made a simple rig sandwiching a magnet
between a small piece of wood and a layer of cardboard taped
together. I marked it that if a magnet was attracted, the magnet
was correct for the block, if it repelled, it was right for an
accessory. This was cumbersome, and it was so easy and common for
a magnet to slip once it was removed from the stack of magnets, forcing
you to start
over.
I then discovered that I had a magnetic catch designed for a
necklace. Its
magnets were slightly larger than the 4mm magnets I was using, and
obviously the two halves had opposite polarities. I labeled one
half "block" and the other "acc" and I was in business. To use
this tool, I would use the appropriate half for the place I was
installing magnets. I would lay out several magnets on my table a
couple inches apart, so they wouldn't fly together, and started
applying my adhesive. To handle the magnet I would move the tool
over a magnet which would jump up and attach itself with the correct
polarity. I would then apply more glue to the magnet and put it
in place, using the tool as a handle. I would then slide the tool
off sideways and wipe any excess glue from the surface around the
magnet. This process went smoothly and quickly, and I never got a
magnet installed upside-down.
Here is an example of installing the very tiny magnet.
Printing:
I started by going through the print list in the
manual and assigning the color I wanted for each piece. I then
started printing all the silver color parts
(as defined at that time) There were quite a few. I was
really pleased with the results of printing with this filament. I
bought it from Amazon, choosing it by the pictures and
descriptions. Descriptions are often
blatantly over rated, and pictures can be deceiving, but I took the
chance. I could not be happier
with the results! The parts look like they are really made of
polished
aluminum! The filament I bought was Mchyi Silk Silver PLA.
A friend bought the same brand in Silk Red, and the results looked like
red anodized aluminum, almost a candy apple red.
After printing a couple small samples to check out the filament, I
printed one of the cylinder heads! I was thrilled with the
outcome!
I then proceeded to print the balance of the silver parts (as initially
defined). The top row contains all the cam lobes and bearing
blocks.
The second row has the valve lifters, the third the pistons, then the
generator bracket and fan, the fuel pump, and the distributer.
Next
come the carburetors and air filter top and bottoms, then the intake
manifolds -lower and upper, followed by the heads. The pen
is included to give scale to the items.
Next I decided since I had to print the block sometime; it might as
well be now.
The first picture shows the block 42 minutes in at the completion of
the first
layer. You can see just how close to the maximum capacity of my
printer it came.
The second picture shows the progress after 24 hours, the third, after
48 hours.
In the last 2 pictures, you can clearly see the infill pattern,
which is only 15% dense, while it is solid at all the surfaces.
Finally, after over 66 hours, the block was complete! I was, and
am, totally happy and amazed with the results! (and VERY relieved that
there were
no "glitches"!
I then proceeded to print all the
remaining parts, one color at a time. I usually printed a number
of different parts at a time, up to 20, to reduce the total number of
printing cycles, although that did not really reduce the total printing
time. In all, I printed 36 loads of parts, not counting the parts
I printed over to change color.
After I finished printing all the required parts, in several different
colors of filament, I
staged the parts and took this "group" shot. Altogether I printed
in silver, gray,
black,
red, and blue. Even thought I generally was going for
realism in my color choices, I went with Eric on the valve
colors. Making them blue and red makes it so much
easier to recognize which are the exhaust and the intake valves when
demonstrating.
The red timing cover and water pump bodies and the black exhaust
headers were
later reprinted in silver. The final total print count was 210.
After
tearing down the setup, I discovered that 3 parts had been missing from
the shot. I located the parts and photographed them separately on
the same
table. Using
Photoshop Elements, I added the 3 parts to the bottom, did some touch
up and got
the above picture of all the parts.
Assembly:
OK, I have all the parts needed, now the delicate and and
sometimes repetitive part starts. I need to put them
together! Fortunately, Eric did a great job writing a detailed
manual with a number of good hints about specific parts!
The first item I assembled was the camshaft. I had printed
20 separate parts
consisting of 16 cam lobes 3 bearing mounts, and one spacer.
These were assembled by stacking them over a 3 mm threaded rod and a 2
mm smooth rod running through alignment holes in each part.
Adding
the cam gear completed the assembly. I first assembled the 20
pieces dry, clamped them between nuts on the threaded rod and took some
measurements. The manual specifies what the assembled dimensions
should be for the front half, the rear half, and the total
length. All of my dimensions were significantly over. I
calculated that each element was, on average, about 0.12 to 0.15mm
(.005 to .006 inches) oversize.
I started trying to sand down individual parts without much
success. I then had a flash of inspiration: why should I try
reworking them when I can re-print them with very little effort.
I
calculated the percentage of their current size I needed to hit the
dimensions (if I recall, it was 98.7%), set the Z scaling for that
scale factor and re-printed them. I re-stacked them, and the
dimensions were right on! Next I took them apart and re-assembled
them one piece at a time, gluing each piece to the previous part, and
adding ball bearings as required. When I
finished and added the cam gear, I had an amazingly realistic camshaft!
These are the parts I had to work with. Note the larger hole in
the center for the threaded rod, and the smaller, off-center hole for
alignment.
Each cam lobe has an identifying number printed into one side to
prevent mis-positioning it.
And here is the completed camshaft. Is it a model, or is it the
real thing?
This is the completed and aligned crankshaft, less the flywheel.
Every moving surface on the crank uses a pair of ball bearings.
Each
connecting rod and all 3 main bearings each have two.
Now we are building an
engine! The crankshaft laid nicely in the bottom of the block and
was secured by the 3 main bearing caps. The pistons and
connecting rods were then assembled. Next the pistons were
installed into the engine just as in a real engine, except there is no
need to compress the rings. Each connecting rod clamps securely
to a pair of ball bearings on the crankshaft using its rod bearing cap
and screws threaded into brass inserts. Rotating the crank now
gives the satisfying sound of the pistons sliding in the cylinders and
you can see them operating in sequence.
The pistons and rods are assembled and ready to be installed into the
engine.
The crankshaft is installed and all the pistons and rods are in place.
The valves were probably the most
tedious portion of this build. Each valve is first assembled with
a valve guide on a cut-down socket head cap screw. The screw
makes its own threads into the valve. Next a lifter with a hex
head screw threaded into its top is inserted into the block, the valve
and guide
are dropped in place and the clearance is checked (with the cam 180
degrees from the lobe). By trial and error, the lifter is
adjusted until there is essentially no clearance, but no pressure on
the lifter. Now the valve has the spring added, held in place by
a retainer. This assembly is placed into the block and the guide
is worked into location (I used a small pointed bamboo stick between
the raised valve and its seat) and locked in place with a
setscrew. The setscrews are threaded into holes in the
block. The intake ones are just slightly deeper than the
setscrew, but the exhaust are very deep and require about a hundred
turns to insert them into position. Thank you Mr. electric
screwdriver! Once that valve is secured in
place you are ready to tackle the next one. I felt a real
sense of satisfaction when I finished installing the final valve!
We now have a pretty much completed engine! (short block) Turning
the
crankshaft causes the pistons to go up and down and the
valves to open
and close in the correct timing sequence for the 4 cycle engine to
theoretically run.
My temporary engine stand can be seen in this picture. It made
most
of the operations on the engine so much easier than if I did not have
it.
The remaining parts of the build
were
fairly routine. The engine pan required 8 magnets to be glued in
place along with a number of setscrews and nuts to simulate it being
bolted in
place
The two water pumps, the fan and the generator all had pulleys mounted
on a pair of ball bearings. I made several simple modifications
to stiffen and remove play from these, but they all went together
without problem. Once they were in place, a pair of black rubber
bands became the v-belts to drive all the accessories. The
distributer is bolted to the timing cover, and the
distributor cap is attached to it with 4 magnets in each part.
This shows the water pumps, the distributor, and its cap. Each of
the cap's wire ports has a magnet glued into its end.
The wiring harnesses were by far the most challenging pieces
remaining. The hardware kit came with 7 feet of 12 ga. solid
commercial home wire. It is very stiff and has identifying
printing on it that would be very distracting. I hoped it could
be removed, but feared it might be printed under the tough clear outer
coating. I tried some rubbing alcohol with no results, then I
tried some some acetone. One quick wipe and the printing totally
disappeared! What a relief!
Each harness consists of 4 lengths of wire running from the distributor
cap to a spark plug. Each end has a printed boot with the
distributor end containing a magnet to attach to a mating magnet in the
distributor cap. The spark plug end has a hollow boot that drops
over the
spark plug. There are several wire retaining holders to keep
things neat. The challenge is to bend the wires just right so
that they match at both the distributor and the spark plug ends with no
tension. They should also look neat! Once the wires are all
tweaked just right, drops of
superglue at the base of the boots and at the wire guides keep
everything in perfect alignment. They were a challenge, but they
came out looking quite good!
These are the spark plug harness wires as initially cut to
length. A light wipe with acetone cleanly removed all the
printing. I then spent
the time to carefully straighten the wires before starting to form them
into position.
The finished harnesses look far better than I could have imagined as I
was trying to bend them into shape. A drop of superglue
every place the wire touches plastic holds them securely in place.
The hardware kit came with a 12 volt gearhead motor which assumes
the role of the starter motor. It also came with a speed
controller, a wall type power supply and several connectors.
The final mechanical part of the build was mounting the gearhead motor
with its drive gear to its mounting block and then attaching that at
the corner of the oil pan to the block.
And here is the completed engine!
Control Box:
The hardware kit came with a wall plug-in type power supply, a speed
controller board, and several connectors to power the motor. I
wanted something a
little more elaborate. On Thingiverse I found a box designed by
Casey Clark (screen name twistclark), specifically to hold the variable
speed controller from this
engine. It is in the shape of a 6 volt battery which adds a nice
touch. Thank you Casey!
In addition to running from the wall power supply, I would
like to be able to run it from a portable power pack designed to
recharge cell phones. The only problem is that these are all 5
volt and use a USB type connector. I have a boost module left
over from a previous project that will output a higher voltage than the
input. It was easy to adjust it to convert 5 volts DC to 12
volts DC. It can
supply much higher current than the engine requires and is a perfect
fit.
I originally printed the battery box as designed, and my speed
controller fit perfectly. I then started experimenting with
adding the boost module and additional connectors and ended up
designing a box from scratch the exact size of the original,
so the battery shaped lid will still fit. The new box has
mounting provisions for 3 circuit boards: the speed controller, the
boost converter, and a tiny board, approximately 1/2 inch square, which
mounts a micro USB connector and provides easy mounting and
connecting. The box also holds an on/off switch and 3 connectors.
I drew up a connection diagram, and the only components needed are the
3 boards, 3 input and output connectors, a power switch, and a
diode. The diode prevents the wall power supply from back-feeding
power into the output of the boost converter.
After I completed the box and everything worked functionally, I was not
satisfied with
the lid which merely sat passively on the box, so of course, I attached
a pair of magnets to the box and another pair to the cover. To do
this, I printed a pair of 6mm cubes with a magnet socket in one
side. I glued these inside the box at two corners the correct
distance down from the top. A pair of magnets surface glued to
the cover completed the mod.
The redesigned base box holds all the additional components I
desired. On the left, the front of the box has a power switch and
the speed
control knob. In the center, the rear contains the 12 volt
connector, a 5 volt micro USB below it, and a motor power connector on
the right.
(That is a much more elaborate connector than required, but it is what
I had.)
The picture on the right shows the added magnets.
It's a tight fit inside the box, but by mounting the controller low and
the boost module high it fit nicely. On the left rear, near the base of
the box is the tiny USB
connector board. The upper right is the supplied speed
controller, and the near side is the boost converter.
A late
discovery:
The latest improvement was created by a
discovery I should have made earlier. During the build, I kept
thinking that the raw drive motor really needed something to disguise
it, but put off any action till later. Then in looking again at
Eric's model
I noticed that he had a cover over the motor. I looked carefully
at the list of files supplied, and there was a starter cover. I
had missed it completely as it was not on the Printed Parts List in the
manual, and it did not jump out at me from the list of the 84 supplied
files.
Needless to say I immediately printed it and it is really nice.
Not only does it cover the bright metal motor, but in the replica
starter solenoid is room for the extra connector supplied in the
kit. This not only makes the general appearance much better, but
it eliminates any hanging wires until you plug in a cable to run
it. The cover is long enough that I could leave enough wire to
make easy connections and the excess easily fit in the end.
The picture on the left shows what the uncovered motor looks like, as
well as the ease of wiring the motor and connector. The right
shows the cover in place
and the connector plugged in to power the motor. I used a turn of
electrical tape on the connector and two layers on the motor and all
the parts
pressed
together quite firmly.