Teething problems '34-36

Several HM-14s fell out of the sky. The accidents were caused by some design mistakes. Mignet placed the pilot behind the center of gravity (CG) and if a heavier pilot (Mignet being a bit lighter than the common person) flew the airplane the nose couldn't be kept up after a dive of 15°. You can see the problem Henri Mignet had in one of the drawings above (the one about shifting lift forces). In that drawing you see that the lift forces can shift backwards. If they do you need to have enough weight behind the CG to have a good chance to counter this moment with the front wing. If the front wing has limited rotation and the weight behind the CG is less than the design weight ... you have lost of problems. Several pilots dove to their death. Mignet quickly redesigned the HM-14 in 1936 a controllable rear wing. Mignet created a nose-up with this control.

The newly controllable rear wing rotated automatically to a lesser angle of attack if the stick was pulled to its extremes. This way the rear wing creates less lift and the moment around the CG gets smaller. The lift of the front wing will now be large enough to counter the moment around the CG and prevent the unstoppable diving, which killed several. The system worked. But already many countries grounded the Flying Fleas. And quickly the Flea fever died. But this didn't stop Mignet. He kept designing tandems.

I quote the book "The design of the aeroplane" by Darrol Stinton. He gives a good and complete view on the possible mistakes in the HM-14 design and how they were solved.

"Four factors appear to have contributed to the HM-14 accidents, and they are worth recounting:

1. The wings had a section invented by Mignet, with a sharp leading edge. Control in pitch was achieved by changing the angle of incidence of the foreplane by a direct linkage with the stick. Maximum incidence was limited to only about 4.8 deg on a specimen tested in the 24 ft (7.3 m) wind tunnel at the Royal Aircraft Establishment, Farnborough (ref 4.19). It is possible that the pitch control lacked authority, and that the lack would have been worsened by the tendency of a sharp leading edge to force premature separation and loss of lift at moderate angles of attack.

2. The foreplane trailing edge overlapped the leading edge of the rearplane. With the foreplane at its maximum incedence, a venturi effect might have been induced in the gap between the planes, increasing the lift of the rearplane, so that the nose-down moment of the rearplane about the centre of gravity could have become larger than the foreplane could counter.

3. Longitudinal stability was dependent upon lift coeffiecient, being most stable at large angle of attack and high Cl, and least stable at low (due to the neutral point varying considerably with the angle of attack).

4. Inadequate control of the centre of gravity. The Farnborough tests showed the aeroplane to be unstable in normal flight with the CG further aft than 0.4 times the foreplane chord. In a dive steeper than 15 degrees recovery could not be achieved.

There is also evidence from later variants that the control power of the foreplane is affected by propeller diameter. ......... A tip that is not high as the foreplane can cause a reversal of circulation beneath the centre section when power is applied, making it harder to lift the nose. A propeller should be about 4 in (10 cm) higher than the foreplane to maintain the authority of the control in pitch. Although Flying Flea variants can fly with quite low power engines, the small propellers used with such engines on gyrocopters, for example, should be avoided at all costs.

1. Among the cures applied to later variants of the HM-14 were:

2. The wing section was replaced by a tried design with a rounded leading edge.

3. The distance between the foreplane and rearplane was increased, so that the overlap disappeared, being replaced with a gap.

4. The rearmost position of the centre of gravity was limited to 0.25 times the total distance between the leading edge of the foreplane and the trailing edge of the rearplane.

Mignet's wing arrangement was close to that of a biplane with considerable stagger and a small gap. Or both wings could be taken as a single low aspect ratio wing, incorporating a slot, in which case aspect ratios varied between 3.3 and 3.43 depending upon the span of the foreplane. Zimmerman showed in 1932 that wings of very low aspect ratio have effective spans about 5 per cent longer than the actual span because of the closeness of the tip vortices, which dominate the aerodynamic picture. .............. It is possible that the unusual, but effective, flying characteristics of the Flying Flea owe something to the phenomenon."

One could rephrase the last chapter by saying that most problems were gone by the use of autostable airfoils. These airfoils have a lift force which doesn't move (too much) when having a larger incidence angle. So there is no more problem with lift forces shifting backwards and increasing the moment around the CG. The moment can no longer go crazy like in the earlier HM-14.

Paul Pontois, a known name in the Pou world, writes that a Pou du Ciel with a tested AUTOSTABLE airfoil, enough frontwing rotation angle (no negative incidence!) and a good positioned CG does not need a Cosandey flap. I found a picture of a glider of Louis Cosandey. The glider had a flap on the rear wing. It was probably used only to move upwards to reduce the lift force of the rear wing and to reduce the moment that killed several pilots. It also is a good glider ratio control. It is my opinion that the rotating rear wing of Henri Mignet had the same function. So if Paul Pontois says that a Cosandey flap is not needed in his named conditions, a rotating rear wing is not needed too...when you use a autostable airfoil.