I’m sure the Teleflex cables used on the Max family work perfectly well, but I would rather fly with a pushrod system for pitch & roll. I just reckon to get a nicer feel, plus it's more inspectable.
And I want to incorporate a 6 degree up setting for both ailerons simultaneously (reflex), to improve fuel consumption at cruise.
I also thought it might be interesting to have the option of using a bit of flap to shorten takeoff & maybe landing roll. Seems a shame to have full length ailerons and not exploit fully their potential.
Anyway, I enjoy a bit of design & build exercise. So I went for my own system of controls.
So, firstly the pitch control. I wanted to make all the elements in this very stiff to give good feel and reduce the chance of buckling under extreme inputs (probably academic if you need that much force!). So I used 2 lengths of straight tubing rather than use any bends.
Because a straight line from joystick to elevator would pass through my groin area when seated, I used an idler to allow the change of angle from cockpit to tail. The idler also allows me to separate the twisting motion of the stick needed for roll from the fore-and-aft movement for pitch.
So I mounted the idler under the seat & made the connection with the forward section of pushrod from the stick by means of a rod end joint. I made this a female fitting to minimize failure in bending more likely to occur in the male bearings. In all of the control circuit ‘joints’, where there is movement of one item relative to another, I used bearings of either rod ends (ball joints) or phosphor bronze Oilite bushes, rather than just drilling a hole through the aluminium. I also do not use the bolt shaft as a bearing surface.
The forward end of the pushrod also connects to the stick via a rod end to allow for lateral movement as well as fore-and-aft. The pushrod passes through a cutout in the modified aileron quadrant, such that there are no fouls at all extremes of input.
The rear pushrod is a mighty 1.25” dia 0.048” tube. I wanted that dia (for Euler buckling purposes) but not that heavy a gauge. Sadly, in this country, you are sometimes forced to take what is available. Pushrod is still lighter than the same length of Teleflex cable though.
I used a plastic support halfway along the fuse to discourage any droop (& prevent buckling), and ran the pushrod out through the rear bulkhead to the elevator. Here I used a second cleat to transmit the input, using 3/8” dia bushes, with 3/8” .058” steel tube as the axle, with an AN4 bolt passing through the whole.
There was a nearly clear path right through the rear fuse, but I did have to steer the aft diagonal fuse brace around the pushrod. The cutout in the station 4 bulkhead clears the pushrod.
So the pitch control was straightforward, and could be used with cable ailerons for those concerned about Teleflex failure in pitch.
The roll control was trickier to design, because of what it has to do. I needed to have a system that gave me asymmetrical aileron movement as per original (up 22 deg, down 16); 6 degrees of reflex plus a nominal 10 degrees of flap. And again, I wanted it to be structurally stiff to feel sweet in the air.
I had to redesign the aileron quadrant plate because the elevator pushrod passes through it. That meant taking the quad pushrods off the top of the plate, which loses the aileron differential.
But to turn the system through 90 degrees, I needed a bellcrank, so this is where I put the differential back in. It works on the same principle: effective & ineffective portions of the arc.
I want to keep my Max as clean as possible aerodynamically, so decided to run the second aileron pushrods inside the cockpit. Downside is sitting next to a couple of moving primary controls, so a cover will be needed. Second upside is I could keep the pushrods very nearly straight, with just a double kink at the end where it exits the fuselage. This makes them stiffer.
The aileron pushrods are 5/8” dia .058” 2024-T3, which is quite hard to bend. But that’s good, because it makes them much stiffer, and stiffer is good (blush). I used some curved blocks fixed to the worktable to form them by hand.
I needed to make the inserts for the end of the aileron pushrods. I wanted to use 2024 aluminium, tapped to take the rod ends, but sadly I don’t have any metal working machinery. I’d love a small lathe & ideally a milling machine, but The Management says the budget won’t stand it
Then I realized my drill press was a lathe, just vertical instead of horizontal. So I found an old bastard file, and ground it into a lathe tool profile. The 2024 stock just fitted in the drill chuck, so I was able to turn down the bar to make a press fit insert.
I tapped the end 10-32, and for peace of mind fixed it with a tapped M4 screw in addition to the tight press fit.
To prove my theoretical design would actually give the correct differential etc, I needed to mock up the ailerons. I did this by making a ply end rib, with the spar pickup and aileron pivot positions accurate. I fitted this to the side of the fuse, then used the aileron pivot point to mount a dummy aileron. I graduated the arc of the aileron in marker pen, & used the ‘production’ horn to drive the aileron. I did this both sides.
So the roll control all worked smoothly & so far as drawn. Next are the reflex & flap settings.
I needed a hardpoint under the seat to mount the aileron bellcrank, so I mounted this on a 1.25 dia ally tube (elevator pushrod spare). This torque tube is pivoted lower down to increase the arc of travel which dictates the reflex & flap deflections.
The torque tube is operated by a handle made from 2” x 2” x 1/8” 6061 angle. The handle doubles as the pivot point. A smaller version, without the handle, goes on the opposite end.
At this point I climbed back into the cockpit to decide on the most ergonomic position for the flap lever. I’d already decided I wanted the throttle control on the right hand wall, using my preferred left hand for the stick. I also brought the stick back 1.5” from the standard position, as that felt more comfortable.
I put the flap lever where it came readily to hand, and drilled through the cleats to drive the torque tube. Then I made a simple 3-position detent, using a pin made from a stainless steel screw, which engages in a rail fixed to the cockpit wall. To disengage, you push the handle outboard, move to next position, and allow the spring in the lever to pop the pin into the next hole. It has quite a positive feel. Labels will come later.
This system gives the required positions for normal, reflex, & flaps as proved by my dummy aileron markings. But an inevitable result is that available aileron movement is reduced in reflex & flap position. But there’s almost full deflection in the reflex position, and while the range is much reduced in the flap setting (the ailerons already using 10 degrees of droop out of the available 16 degrees downward deflection), I think it will be adequate. I don’t expect to use the flaps in any sort of crosswind, and prefer sideslips for shortfield landings anyway. But the flaps might be useful, particularly for takeoff from a bush strip.
Lastly, I made some covers to protect the aileron pushrods. You can see the guide tube for the rudder cables under – I routed those cables under the seat to keep them away from the sides.
Oh, and I heard some scruffling from my shop logburner. Opened the door & found a furry visitor.
My internet connection is sooooo slooow I'll have to stop - maybe post some more later.
Thanks guys for all your kind words. I'll let you know how it goes in the air . . . in about 6 months. I'm definitely into the "90% finished, 90% to go" period.
this bell-crank is already almost in straight line, if with full stick to the side it flips over the center to the other side since there is nothing to prevent it as it looks could happen very easily you are a dead man.
this bell-crank is already almost in straight line, if with full stick to the side it flips over the center to the other side since there is nothing to prevent it as it looks could happen very easily you are a dead man.
I think George has made a very astute and scary observation. I'd check full/extreme range of motion possibilities before flying this... As others have said, your work is beautiful; but, I would double/triple check design of this setup. I look forward to learning what you discover.
Why focus on proving how great you are, when you could focus on becoming better?...
Good eye George, that would be a disaster. Bruce could you lengthen the center tube and move the bellcrank farther down to change the angle of the pushrods in reference to the outer bellcranks?
Sorry, Phil – I missed your post; & thanks to TTT for reminding me (& for your post).
Phil asked “What if the idler's wooden block failed?” There’s two answers to this, one short, one long. Short answer first: Almost nothing would happen. And here’s my thinking on why (cue long answer):
Firstly, we have to consider possible modes of failure of the wooden block. These might be in tension (block is pulled up off the floor); compression (block is forced through cockpit floor – though this is more of a floor failure); pivot bolt pulls out of block. I can’t think of any more – there are negligible lateral loads on the block.
So how is the block loaded? Well, in normal use, there are only vertical loads acting on it. The geometry of the system is such that the idler lever travels through a very short arc, with the forces at the top acting nearly tangentially. Ignoring the very low frictional forces in the block bearing, this means the loads are close to vertical, down (or up) through the idler lever. It is not possible to apply the load other than in this direction.
So let’s quantify the load. In the diagram below, the upper drawing shows the arrangement of the 3 elements at the idler. Let us assume the maximum force we’re able to pull generate on the elevator is 20lbf. (This is a hell of a load if you consider your normal inputs in flight, & anyway, it’s just an OTT stab at what might be possible). So for equilibrium, the load in both forward & aft pushrods is 20lbf. Because of the change of direction – the 11 degree angle between the 2 p/rods – there is a reaction at the idler arms, and hence at the wooden block in question.
This force is resolved graphically in the lower diagram. The two 20lbf loadings & their direction are drawn as vectors (i.e. a line showing both force & direction, drawn to scale) and the force in the idler arms is the resultant vector (the line that closes the triangle of forces).
So from measuring this closing vector we find that for a 20lb force on the pushrod the idler arms see 4.5lb of reactive force. See attached diagram. To this load must be added the 55% or so of the weight of the two pushrods. This is roughly another 2lb.
Under this load case that means a force of 6.5 lbs is pushing down on the wooden block. That is roughly just over 4 mugs of coffee sitting on the block: hardly enough to burst through the cockpit floor or split the wooden block asunder And nowhere near enough to tear the ¼” dia pivot bolt from its housing in the solid oak wooden block. If you pull a 20lbf load (instead of pushing), the value would be the same (less the 2lb weight giving 2.5lbf), but the force will now act upwards. Same comments apply, but epoxy is good for better than the approx 8oz per square inch bonding strength required to resist the loads trying to pull the block off the floor.
I suppose it’s just conceivable that one of the phosphor bronze bearings may seize solid – though extremely unlikely! If this were to happen, then the 20lb force applied at the end of a 3.5” lever would give a 70lb moment trying to peel the block off the floor – again, nowhere near enough to fail either glue or wood.
Anyone still reading, or back from the popcorn run, may recall I said earlier that even if the block did fail, almost nothing would happen. The reason for this is simple: the aft pushrod will be located by the frame 4 bulkhead through which it passes. The attached photo shows the clearance round the pushrod is minimal, so the pushrod would be well supported, and the system would operate pretty much without any noticeable effect.
So, hope this answers your queries. I tried to use words rather than calculations, but others may wish to run the numbers if they are really really stuck for entertainment.
I'll answer the question re the weight of my system versus stock when I've done a bit more work in the shed - there's only so much time I want to spend over a hot keyboard. . .
Logged
