My powerboat building spree kindled an ongoing fascination with propellers of all kinds. When interest finally turned to aeronautical props, I channeled it into (i) a series of motorized prop-carts (MPCs) driven by pusher props, starting with MCP1 posted several months ago, and (ii) a working horizontal-axis wind turbine (MOCpage coming soon).
∧ MPC2, the MOC presented here, is a 2nd-generation MPC design with more than double the top speed of the MPC1 prototype. MPC2 is also much easier to control and powers right through small surface irregularities (like the grout lines in the photos) that routinely stopped MPC1 dead in its tracks. I think it looks a heck of a lot better, too.
∨ The video below shows MCP2 turning laps with its original 2-blade prop.
Early on in MPC1 development, it became clear that for a LEGOŽ prop-cart to perform well at all, the following interrelated and somewhat contradictory design goals would have to be met:
A motor-gearbox-prop (MGP) combination carefully optimized for maximum thrust.
High prop speed.
Lightweight prop construction capable of withstanding the high centrifugal stresses involved with a wide margin of safety.
Minimum bearing and rolling friction.
High structural rigidity.
These goals remain valid. As I learned only while photographing MCP2 for this page, however, correctly identifying an MOC's key design goals and understanding -- and acting on -- their true order of importance are 2 very different things.
When MCP2's much larger prop failed to yield the expected performance gain, I knew something else was holding it back but couldn't find the real culprit. Had I not noticed the badly sagging rear axle in the 1st batch of photos I took for this page, I might still be scratching my head. For an axle support fanatic like me, that's pretty embarrassing.
The good news: That fix was a heck of a lot easier than wringing more thrust of the prop.
There's something to said for starting a MOCpage before all the wrinkles have been ironed out. A fresh look at an MOC with unresolved issues -- e.g., through the lens of a camera or the eyes of a friend -- often brings the hidden or unsuspected causes to light. Writing about a problem MOC is even more potent in this regard.
Keep an open mind as to what's most important. Correctly identifying an MOC's key design goals is one thing. Understanding their true order of importance is another.
Any MOC reliant on a prop for propulsion (e.g., a prop-cart or powerboat) or on a rotor for power (e.g., a wind turbine) will benefit from a careful matching of motor/generator to gearbox to prop/rotor to external load based on testing, preferably quantitative.
∨ The next 8 photos provide an overview of MPC2 fitted with its original 292 mm, 2-blade prop geared up 1:9. The cosmetics are greatly improved over those of the bare-bones MPC1.
∧ Two 11x5 Technic frames joined by 1x9 liftarms form the structural core of the strong but light chassis.
∧ View from above gives an idea of MCP2's overall size. Light-colored floor tiles were 305 mm (12 in) squares before corner cuts.
Rear axle support fiasco -- a cautionary tale
∨ Problem: The photo below shows my first attempt at jacking up MCP2's heavy rear end to accommodate a larger prop.
The prominent sag in the one-piece 12L rear axle indicates a serious shortage of axle support, but I noticed it only on seeing this photo some 2 weeks later. When the larger 292 mm prop failed to produce the expected performance gain, I knew something was awry but couldn't find the problem -- until I saw the photo above.
I never thought to take a close look at the rear axle under load -- a rather embarrassing lapse for an axle support fanatic like me. Not quite sure how that happened, but I'd been obsessed with new designs for the larger prop for weeks, and -- duh! -- MCP2 seemed to roll freely with a manual push.
As with most of my MOCs, MPC1 and MPC2 have always operated at the very edge of what LEGOŽ motors, parts, and materials can do -- a setup for just the kind of fiasco I'm describing here. MPC2's new 292 mm prop turned out to be huge successes once it was truly free to roll.
The excessive bearing friction and other losses caused by the axle sag were all it took to hold MCP2 back against the added thrust of the powerful new prop. In retrospect, MPC1 performance had suffered from the same problem, though not quite to this extent.
∧ Solution: Splitting the rear axle to eliminate binding in turns and adding outboard axle supports eliminated the excess bearing friction induced by the badly sagging one-peice rear axle. Of all the changes leading from MPC1 to MPC2, this fix had the greatest positive impact on MPC2 performance -- despite the added weight.
Note to self: Never underestimate the trouble a poorly supported axle can cause under load.
Motor-gearbox-prop (MGP) optimization
∨ MPC2 drivetrain: Removing the nacelle covers reveals the entire MPC2 drivetrain -- from left to right, Power Functions (PF) V2 IR receiver, PF L motor, 2-stage overdrive gearbox, and original 292 mm prop.
∧ When this photo was taken, the gearbox still had MPC1's overdrive ratio of 1:15.
∧ The overdrive was later changed to 1:9 to better match MCP2's larger prop to the L motor's power-RPM curve. Note the generous prop axle support.
The 7.4V PF Li polymer rechargeable battery is a no-brainer for a weight-sensitive MPC. It's the lightest and longest-running PF battery available by a wide margin, suffers the least voltage sag under draw, and delivers more electrical power than 7.2V PF AA and AAA battery boxes filled with 1.2V NiMH rechargeables.
The L motor emerged early on as the best PF motor for a LEGOŽ MPC. The higher-power pre-PF RC Race Buggy motor used by MPC3 turns out to be even better in this setting, but the L motor is still an excellent choice.
Like all DC electric motors, the L delivers maximum mechanical power when loaded down to ~50% of its no-load speed (NLS) and maximum torque when loaded to a stall. At the 7.4V delivered by the Li battery, the L motor's NLS is ~300 RPM. In an MCP with a given prop, the motor's torque and mechanical power outputs determine prop spin-up time and thrust, respectively.
Since the ultimate goal of optimization is to maximize MPC top speed by maximizing thrust, the L motor has to run at peak mechanical power output. That gives us a straightforward guide to MGP optimization for DC motors requiring nothing more than a cheap tachometer to follow:
The target motor shaft speed for any MGP optimization is half the no-load speed.
It's that simple, and my anemometer readings have backed it up without exception. When the MGP optimization involves an L motor powered by the PF rechargeable (as with MPC1 and MPC2), the target shaft speed is NLS / 2 = ~300 / 2 = ~150 RPM. Every prop I've tested with that setup has produced its highest peak downstream air speed when the gearing resulted in an L motor shaft speed close to 150 RPM.
As with MPC1, objective measures of prop performance were needed for the long, iterative process of motor-gearbox-prop (MGP) matching.
∧ Tools of the trade: Handheld laser tachometer on the left, handheld anemometer on the right. Both are inexpensive. As detailed here, prop speed is easily determined with the tachometer. The corresponding motor shaft speed can then be calculated from the known overdrive ratio.
The anemometer isn't really necessary for the matching process, but it does provide a quantitative measure of prop output via downstream air speed, a crude index of thrust. This measurement is greatly complicated by the fact that prop slipstreams are never uniform and necessarily include a great deal of swirl.
However, slipstream speed measurements can still be made meaningful by sticking to an arbitrary reference distance downstream from the prop (here, 0.30 m) and using the anemometer to find the fastest part of the slipstream there for every MGP combo tested.
Original 2-blade prop construction
∨ Building safe, effective high-speed LEGOŽ air propellers turns out to be quite challenging. Eye protection, patience, and endless testing of parameter combinations are essential. A willingness to resort to glue in the name of safety also comes in handy.
The original 292 mm, 2-blade MPC2 prop below felt rickety but stiffened up nicely under centrifugal force and managed to stay in one piece at its operating speed of 1,309 RPM without glue.
∧ Each blade is mounted on a 4L axle with a stop at the blade end.
∧ The paired 1x2 cross-blocks at the outer end of the 1x7 liftarm serving as a prop strut seize against each other and the 4L blade axle when centrifugal force tries to pull the blade free.
∧ The sandwich-style hub provides the necessary central axle hole for the prop shaft. Technic plates with closed ends would have made for a stronger hub, but they don't come in yellow, and this hub never showed any inclination to come apart.
The main parameters to be juggled when designing and building a prop for use in air are (i) prop diameter, (ii) blade angle, (ii) individual blade area, (iii) number of blades and total blade area, (iv) total mass and rotational inertia (aka moment of inertia) about the prop shaft, and (v) the target tip speed the prop must reliably sustain without ending up as a source of shrapnel.
Each of these interrelated parameters affects both prop performance and strength -- often in ways that are very hard to predict. That's where the testing comes in.
∨ MPC2 saw a 10-15% top speed gain when the 292 mm, 2-blade prop shown above was replaced the lighter 340 mm, 3-blade prop below.
∧ NB: All joins susceptible to failure in tension under centrifugal force had to be glued to keep this 3-blade prop from flying apart above a few hundred RPM. Eye protection was an absolute must when testing it prior to gluing.
Since MCP2 owes much of its success to its prop, the choice of blades merits further comment.
The 2 main aerodynamic forces acting on wings and prop blades alike are, of course, lift and drag. By definition, drag acts antiparallel to the direction of relative motion between the blade or wing and the surrounding air, while lift acts in a perpendicular direction. With a suitably designed prop, you can generate thrust with either lift or drag, but the upside potential is much, much greater with lift.
At bottom, real prop blades are rotating wings designed to generate thrust via aerodynamic lift. All well-designed real-world blades have the following interrelated attributes:
Lift-generating airfoil-like cross-sections with high lift/drag ratio.
Very smooth surfaces and edges.
Tapering chord length (i.e., straight-line distance between leading and trailing edges) from root to tip.
Sufficient twist -- i.e., decreasing blade angle (i.e., acute angle between chord line and prop plane) from root to tip -- to insure a proper angle of attack at every point along the blade.
The greatest length and least thickness allowed by aerodynamic, material strength, clearance, and other constraints.
Hence, high strength/weight, length/weight, and length/thickness ratios.
The LEGOŽ elements most closely matching this description are (i) the wind turbine blades (89509) found only in the LEGOŽ Education Renewable Energy Add-on Set (9688), and (ii) the mirror-image #5 and #6 long, smooth Technic fairings. These 3 parts have many important aerodynamic advantages over other potential LEGOŽ blade elements for use in air.
∨ The good news: Curved plates also make good airfoils. A thin curved plate forced through the air at a suitably small (5-10°) angle of attack can produce almost as much lift as a classical airfoil of the same length and width.
∧ The more or less end-on views above show a slender, tapering #5 Technic fairing on the left and a fan-like wind turbine blade on the right. Both qualify as curved-plate airfoils. When mounted for optimum angle of attack and run at high speed, the #5 and #6 fairings generate useful thrust via lift. The wind turbine blade produces even more.
All have smooth surfaces and edges; high strength/weight, length/weight, and length/thickness ratios; and inadequate twists. The very expensive wind turbine blades weigh less, have no obstructions to air flow along their lift-generating surfaces, and are much easier to mount securely to a suitable prop hub, but the Technic fairings do surprising well in props designed to take full advantage of them.
∨ A 276 mm, 3-blade prop made with wind turbine blades pushes slightly heavier MPC3 below to a top ~speed of ≥1.5 m/s -- some 25% faster than MPC2's best top speed of >1.2 m/s with its lighter 340 mm, 3-blade prop. In this photo, that 340 mm prop (yellow) was mounted in front of MPC3's white prop for comparison.
∨ The changes incorporated in MCP2's evolutionary front end (next 3 photos) took square aim at MPC1's miserable steering accuracy -- a problem only compounded by the new large, hard plastic wheels added to reduce rolling friction and improve MCP2's ability to overcome small surface irregularities. Unfortunately, only a modest improvement in steerability could be claimed in the end.
∧ MCP2 uses a lighter, stronger, more compact, and more positive version of the simple self-centering design MPC1 borrowed from Sariel's Technic bible (see references).
∧ Upping steering motor reduction from 9:1 to 15:1 improved steering accuracy at the expense of some responsiveness. The net result was positive, but MCP2 is still too hard to drive.
∧ The 6 white elastics continue to earn their keep by taking play out of the system. (Even at 9:1, they had no hope of centering the wheels against the internal friction of the steering M motor.)
As with MPC1, steering's frankly impossible with a standard 3-state (full forward, off, full reverse) PF remote. A PF speed control is an absolute must here -- usually with steering inputs limited to the -2/7 to +2/7 range. I generally avoid the speed control like the plague, but I've managed to make mine a little more tolerable with the modifications shown in the 3 photos below and described in more detail here.
∧ Before and after: The spartan MPC1 (left) was downright ugly. Adding fenders and other smaller splashes of yellow here and there gussied up MCP2 a good bit. Though functionally motivated, the much higher and wider wheels are more in scale with the rest of the vehicle.
∧ The yellow 1x11 diagonal tower brace provided a convenient mount for a #5 Technic fairing serving as left rear fender.
∧ The left nacelle cover featured here consists of #2 and #5 smooth fairings joined end-to-end.
Tables of features and stats
Overall specs and design features
272 x 118 x 276 mm in LxWxH with prop vertical
0.40 kg (0.88 lb)
Studless except for battery and steering motor mounts
Vehicle top speed:
Over 1.2 m/sec (with later 3-blade prop)
Propulsion motor to prop overdrive ratio:
Maximum propulsion motor shaft speed:
145 RPM (48% no-load) at full power
2 in all -- 1 L for prop; 1 M for steering
IR receiver version:
IR receiver connections:
2 -- 1 for each motor
7.4V PF Li polymer rechargable battery
Modified LEGOŽ parts:
None on original prop; glued hub on final prop
Sariel for the steering mechanism; otherwise an entirely original MOC
Propeller specs NB:Data refers to the original 292 mm, 2-blade prop unless otherwise noted.
Centrifugally stiffened 2-blade
Technic #6 long, smooth fairings as blades mounted on thin 1x7 liftarms
Chord at 70% radius:
Blade area, total:
Blade angle at 70% radius:
Pitch at 70% radius:
1,309 RPM at full power (7.4V)
20.0 m/s (44.7 mph) max
Reynolds number at 70% radius:
~6,900 based on chord
Air outflow speed:
~2.5 m/s peak at 0.30 m downstream
Kmiec, Pawel "Sariel", 2013, The Unofficial LEGOŽ Technic Builder's Guide No Starch Press. Sariel's one of the most talented and prolific technical LEGOŽ builders around, period. I can't recommend this book highly enough. IMO, there should be a well-worn copy in every technical LEGOŽ junkie's workshop, regardless of experience and skill level. You can also learn a great deal from Sariel's well-illustrated web site and countless YouTube videos.
WikpediaYou're likely to find a good explanation of anything discussed here on Wikipedia. You can't believe everything you read there, but it's a superb resource for LEGOŽ engineers. In my experience, the science, engineering, and math articles are all very good to excellent. I have yet to find any real misinformation under those headings, and you're unlikely to find a topic that isn't covered.
Wald, Q.R., 2006, The aerodynamics of propellers, Progress in Aerospace Sciences 42:85-128 -- an treatment if you have access and don't mind the heavy math.
NASA's FoilSim III Student Version 1.5a interactive airfoil simulator. Among the available airfoil profiles is a curved plate simulation that sheds some light on the aerodynamic behavior of the Technic #5 and #6 long smooth fairings when used as prop blades.
Wow Jeremy, I am the second one to comment on this one! It has been 2 1/2 years! I found one of your creations that was at the bottom of the popular list! I have some that have not been commented on or even liked!
Quoting Yann (XY EZ)
Wow that's what I call engineering! It looks fantastic, and the level of technic is extremely high here. Keep building! 5/5 8-)
Yann, thanks for the kind words. I deleted this MOCpage shortly after posting it on acquiring MPC3's Race Buggy motor and wind turbine blades. Resurrecting it now to show that you can still get good performance out of an MPC with much a less expensive motor and set of blades.