Hydrofoil Fin

[clang]

required human power input across the cycle, nevermind the spikes - that's model hitting angle limits, not addressed for now

 

[clang]

mechanically coupled CFD numbers are in

the foil movement is not preporgrammed with these, but reacts physically to the flow/spring interactions
it improves peak efficiency - it went up to 87% in the best case, around 85% is now in the band of very much possible

the next refinement pass will address the power input - right now the only preprogrammed thing left is the foil motion (foil reacts to it)
it leaves us with a required power input from the swimmer which isn't very human

next I want to put model human leg power output through the kick phase and hook it into the CFD simulation
thus I'll be able to see what load response we need for a typical human kick to produce greatest thrust/efficiency
and we could tune the design against somewhat realistic human swimming biomechanics

@7BDiver do you maybe know what power human kick produces across the cycle phases? I think I saw you mention something like that
"definitive" foil CFD needs to take biomechanics into consideration here

a sample graph of required power input, don't take the numbers too seriously, it's one of the extreme case simulations:

sudden jumps are where it bumps into the angle range limiters, we definitely need some cushioning there at least, at best - a compliant mechanism with nonlinear response

 

[7BDiver]

I am at a little bit of a loss to where the friction is most important along the amplitude of the foil, I'm guessing at the steepest angle relative to the strut? Is the heaving motion is making the center of lift move forward of the static center of pressure (25% chord) making the nose roll and pitch up creating more lift? If significant, the heaving motion itself could increase the upper limit of effective angle of attack. The CFD is assuming the direction of flow for the foil is coming parallel to the strut/diver, but is it realistically more of an angle between the direction of heaving and direction the diver is heading. In my mind we want to demonstrate trust which is perpendicular to lift in the CFD. How is the CFD for a propeller set up? The angle of the foil in this simulation would simulate a "feathered" prop relative to the direction of flow.

 

[7BDiver]

mechanically coupled CFD numbers are in

the foil movement is not preporgrammed with these, but reacts physically to the flow/spring interactions
it improves peak efficiency - it went up to 87% in the best case, around 85% is now in the band of very much possible

the next refinement pass will address the power input - right now the only preprogrammed thing left is the foil motion (foil reacts to it)
it leaves us with a required power input from the swimmer which isn't very human

next I want to put model human leg power output through the kick phase and hook it into the CFD simulation
thus I'll be able to see what load response we need for a typical human kick to produce greatest thrust/efficiency
and we could tune the design against somewhat realistic human swimming biomechanics

@7BDiver do you maybe know what power human kick produces across the cycle phases? I think I saw you mention something like that
"definitive" foil CFD needs to take biomechanics into consideration here

a sample graph of required power input, don't take the numbers too seriously, it's one of the extreme case simulations: View attachment 61865

sudden jumps are where it bumps into the angle range limiters, we definitely need some cushioning there at least, at best - a compliant mechanism with nonlinear response

I would guess the kicking force to be around 50 lbs/223 N, that would be flexing the fin quite a bit.

 

[clang]

I am at a little bit of a loss to where the friction is most important along the amplitude of the foil, I'm guessing at the steepest angle relative to the strut? Is the heaving motion is making the center of lift move forward of the static center of pressure (25% chord) making the nose roll and pitch up creating more lift? If significant, the heaving motion itself could increase the upper limit of effective angle of attack. The CFD is assuming the direction of flow for the foil is coming parallel to the strut/diver, but is it realistically more of an angle between the direction of heaving and direction the diver is heading. In my mind we want to demonstrate trust which is perpendicular to lift in the CFD. How is the CFD for a propeller set up? The angle of the foil in this simulation would simulate a "feathered" prop relative to the direction of flow.

where the friction is most important along the amplitude of the foil, I'm guessing at the steepest angle relative to the strut?
it did matter most when foil angle was preprogrammed - the cosine foil AoA trajectory needed nonlinear the hinge torque, it had some friction component
the friction works at reversals - it holds blade's deflected state a little bit there, it's was an overall efficiency gain
CFD with physical feedback driving foil's AoA has shown that just a spring gets very close to that cosine AoA pattern

Is the heaving motion is making the center of lift move forward of the static center of pressure (25% chord) making the nose roll and pitch up creating more lift?
across simulations the centre of pressure sits at 0.26-0.4c
unsteady added-mass force acts at roughly 0.5c - midpoint, taking CP back

If significant, the heaving motion itself could increase the upper limit of effective angle of attack.
yeah, it takes the stall angle from theoretical 10-12 degrees to 12-18 degrees of dynamic stall angle in these simulations

The CFD is assuming the direction of flow for the foil is coming parallel to the strut/diver
the setup assumes flow is parallel to the direction of swimming, and the foil moves through it, the foil sees the vector sum, it's angle of attack should be pretty realistic through
the custom part is that in the latest iteration the foil pitch is a solved DOF instead of preprogrammed
I want to take it a bit further and make the motion itself a solved DOF


I might have misunderstood the question, or its parts

and in general I've been posting too much stuff, I should probably do it less frequently


right now the main mechanical engineering concerns I have is how the biomechanics - how to make it work most where the human body exerts most power, and how to make the angle range limiter engagement smoother, ideally the spring would get a stiffer linearly most of the way, but get very stiff very fast close and at the angle range limit

I'll get back when I have the fully passive / elastically-mounted CFD (heave becomes a DOF) - then I could run it against human biomechanics models: where in the kick cycle how much power can a human apply

 

[clang]

in short one blatantly unrealistic thing still left in CFD is that heave is preprogrammed, making it a degree of freedom allows to put biomechanics models into the simulation - heave will be driven by "realistic" power input

right now it assumes power on demand basically for the programmed heave

 

[clang]

I feel like friction will turn out to be not important
but non-linearity of the spring probably will be very important for ergonomics

it's okay to tune the spring for slow-ish swimming, it performs well for fast swimming nonetheless if there are hard angle limiters
but then to not have a sudden wall of resistance at the angle limit I'd love it if the spring was linear most of the way, and accelerated at close to the end of its range of motion basically, that's the non-linearity I want in it

but first I'll work on the CFD model a bit
and I'm very much open for ideas/corrections etc, I'm relearning CFD with this one in a sense

 

[clang]

Are those struts in the original dol fin pilot plastic?
The yellow on top and black on the back, the big pieces connecting wing to the shoe plate
@cdavis, am I understanding it right?

 

[cdavis]

mine are aluminum. Pretty sure the production model was also.