0
Quagmirian

My little project

Recommended Posts

darkwing

Close to the ground is a very risky place to do stall tests on experimental canopies.

Yes, well, I do lots of stuff I shouldn't really do. At least now I know the stall point I can fly more confidently and work to make improvements.

Someone just happened to be walking by when I was launching the other day and took these interesting but not necessarily useful pictures.

[inline small1.jpg]

[inline small2.jpg]

[inline small3.jpg]

[inline small4.jpg]

Share this post


Link to post
Share on other sites
I lost my reply so I'll try again.

I'll toss out a few thoughts of what you might learn from a peace of soft ware like that.

The first is a warning to try and keep in mind exactly what it is and to try and anticipate it's limitations. All models have a point at which they fail. They are all bullshit straight out of some bodies ass. At best you can hope that they have a passing resemblance to reality over some small range.

As an example, the nose cut. I'm guessing that it's seeing it as a flat plate with no flow across the surface. Well that's not true. Even if it were I'd still be a little suspicious of the sharp corners. That's not to say that it's not interesting. It shows very well a preasure gradient across the nose opening. A fairly significant one. You can see how air will flow in through the bottom half of the nose and out through the top. That in it self would fundamentally change the flow around the model so you know there are errors there. But it does give you an idea of how you might be able to change the nose cut or add a lip to the canopy. Ideally the opening would be right at that purple spot near the bottom.

But that is just one AOA. If you look at how that changes as the AOA changes you can get an idea of how much lip you could get away with on a canopy. At a shallower angle you don't want the higher pressure to shift too far above the lip or onto the edge of the top skin. That's where you start to get a dimple and in the most extreme cases a leading edge collapse. Remember that the model is breaking down there and that this feeds back into it self. I'm not saying that it can't tell you things just that you have to keep in mind that it's at best a rough cut.

The tail is another place that I'd expect some error. Remember it wants to inflate and become rounded over most of the span. If I was to take a guess I'd expect a higher drag from this then you see in the model. Same thing at the nose. It will be draggier then this model.

Having said all that I would expect the lift and CP to be about on. Drag should be at least workably close. What it give you is something you could maybe use as a starting point for a finite wing. What you have there is two dimensional. It's actually no where close to the performance of the canopies wing. With the AR as low as it is the drag of the canopy will be dominated by the induced drag of the wing. Our glide angles suck and it's mostly because of the aspect ratio being under three. To really get any idea of the performance or behavior of the canopy you'll have to look at it as a finite wing. Bad news is that the problem is an order of magnitude harder then that little 2D app. I don't know if you can find any free software out there that can really do it. There might be some hope. There are a few old school approximations out there that might give you something. You could try looking at it with some form of "lifting line theory". Basically it looks at the wing as a sheet of vortexes shed by each section of the span of the wing. It's a way of looking at how each of the sections affects the local AOA of all the other sections. You wind up with this big pile of interlocking equations that go into a big matrix... it's not real pretty. But at least under certain circumstances a computer can muddle it's way through it. I want to say that we did program for... I think it was a tapered twisted wing. It's a standard assignment for undergrads but it's been many a year. You might be able to find some thing out there for free that would give you some thing you could work with. Some thing like a vortex lattice would give you an idea of the induced drag but I'm not sure about the aspect ratio. Normally people look at longer wings then this. It's more reasonable to model some thing like that as a line through the ac. A model like that might tend to break down with such a low aspect ratio wing. If you're going to run in to any problem with something like this it would probable be there.

Another issue is the anhedal angle of the wing. If your lines are long enough you might just be able to ignore it. Let's say that at your trim angle at your normal glide that the angle of attack to the free stream is constant along the span of the wing. I think you could dismiss the curvature for an analysis like this and then look at it separately in terms of the loss of lift due to the curve. In other words bend the wing after you analyze it. At this point you should be able to do a stedy state for that AOA. For that AOA it would give you a lift, drag, L/D and CP. That should let you do a very good guess at a trim for a line set that would give you that AOA and glide.

Once you work that out as your steady state trim angle I think you could look at pitching the canopy from that by looking at the local AOA at each section of the wing as the curved wing pitches. You could then treat it as a twisted wing for the analysis at that perturbed angle of attack. In that way I think you could start looking at things like the pitch stiffness of the canopy and maybe some of it's dynamic characteristic's.

What I'm describing is awfully rough and crude and at times a slight missuse of some of these models but I think you could get away with it and that it would at least yield something interesting. It might give you a starting place for the trim of a new canopy. Just keep in mind that it would be a rough first cut.

And what does that logo panel say on the side of your canopy? I can't read it. We need to see that.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
A little update here.

I've changed my design again. I'm going for a 210 sq ft nine cell, made from the same lovely coloured fabric I used for the brown thing.

[inline parts.png]

I got some brass grommets set in my slider, but due to a misunderstanding the inner diameters are only 3/4 of an inch. Is this big enough for a canopy lined with Dacron or will I have to go with Spectra?

Also, here's a diagram which doesn't really show much but will probably convince some people that I know what I'm doing.

[inline 2794_194_139_small.png]

Share this post


Link to post
Share on other sites
That's going to take a little more explanation.

Where exactly are you getting that drawing from?

Where are you getting the glide angle from? Measurement on the last canopy?

Trim the same as the last canopy?

Are you assuming the suspention point from your java foil and what aoa are you assuming?

And then what? you're going to the aoa of the wing as a whole at center point?

I'm just not sure I understand your design path.

And what aoa are you predicting? It looks pretty high from that drawing. Looks rather flat trimmed.

If that is the aoa that you are expecting relative to the center cell of the wing I'd incurage you to try doing a rotation around a line parallel to your glide slope through the y of the riser. If you think you've got a grasp of the aoa and glide of the canopy then that's what the spread sheet I sent you should give you. Right now looking at your panels It looks like you are rotating it parallel to your bottom skin and I think you'll find that that is equivalent to putting a twist in the wing. It wants to twist the corners nose down and you might see some end cell closure or roll under of the top skin from that.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
Quote

That's going to take a little more explanation.

Where exactly are you getting that drawing from?

I drew it myself in PS. I used my new trim and I have also wrongly assumed that each line will be evenly tensioned. Is there any way I can improve this?

Quote

Where are you getting the glide angle from? Measurement on the last canopy?

The glide angle really is just a guess based on my current brown thing. I'm assuming in increase of glide from 2.5 to 3.5, based on a thinner airfoil, higher a/r etc.

Quote

Trim the same as the last canopy?

The trim is also just a guess from the last canopy. That was based on the Pd 7 cell and then flattened. This trim is based on a flattened PD 9 cell.

Quote

Are you assuming the suspention point from your java foil and what aoa are you assuming?

I'm assuming that the suspension point will be at the 25% chord point.

Quote

And then what? you're going to the aoa of the wing as a whole at center point?

I'm just not sure I understand your design path.

And what aoa are you predicting? It looks pretty high from that drawing. Looks rather flat trimmed.

The aoa in full flight looks to be about 11 degrees, which makes it pretty flat trimmed. I do prefer flat canopies though.

Quote

If that is the aoa that you are expecting relative to the center cell of the wing I'd incurage you to try doing a rotation around a line parallel to your glide slope through the y of the riser. If you think you've got a grasp of the aoa and glide of the canopy then that's what the spread sheet I sent you should give you. Right now looking at your panels It looks like you are rotating it parallel to your bottom skin and I think you'll find that that is equivalent to putting a twist in the wing. It wants to twist the corners nose down and you might see some end cell closure or roll under of the top skin from that.

So this will just generally improve performance in all areas then? It sounds like a good idea and I can see where you are coming from.

Share this post


Link to post
Share on other sites
The lines are certainly not evenly tensioned. The lines are supporting the load on the canopy. The load cord wise on the canopy is dependent on the air flow. The java foil program is a classic two dimensional model where they break it down into a bunch of little panels along the surface with 0 flow across that boundary. You can look at a lot of things with this but what is most relivent here is that it can give you a pressure disrtobution in the flow around the wing at the surface. And it should be fairly accerat in that regard. So the the summation of this pressure around the top and bottom of the wing gives you lift, drag, and moment numbers for the airfoil but remember these are numbers for two dimentional flow around the airfoil. Lift should be pretty good, drag will probable be a bit lower then reality, and the moment or location of the center of pressure should be about right.

So for that local AoA the canopy will try to center it self so that you hang beneath the center of pressure. With a positive camber I'll take a guess that it will probable be some where around .3 cord or a bit higher. Often they will talk about forces around the AC or the .25 cord point. But this also includes a moment function around the AC in relation to the angle of attack. This is a way of looking at the movement of the CP as the AOA changes. You might find it easier to just plot the CP relative to AOA.

As to the tension on each line. If you were really interested you could intergrate the pressure along each section of the wing top and bottom and look at the forces on each line. In theory it will affect the movement of the cascades as the AoA changes. but it's probable not worth worrying about at this point. I'd just think of it as being geometricly riged and fixed and look at the pitching as a whole and not worry about what are basically aeroelastic issues. There are actually a lot of interesting questions there about how each section would like to try to shift to it's own AoA and how the inflation of the wing would like to bow the plane form but I would ignore all of that now.

Any increase in AR will improve your performance. I'm not sure where you are getting a change of 2.5 to 3.5 from. There are some basic rule of thumbs for the fudge factor used in the inducided drag equations. a rectangular untwisted wing is a standard example that they will give numbers for if only to show it's inefficentcy. They may even show graphs and data all the way down into the AR you are at. But most people don't build things that low and you may find that it's not that accerate and is just an extension of the graph to give students an idea of why you should never build a wing that short. The next step up would be to look at some form of lifting line theory. You might be able to find some code for a program that will do that for you. As I recall the restaned case of a tapered twisted wing was not too bad. But at these low AR's I wouldn't put too much trust into it till I kited it or flew it.

There is a long tradition of steeling other peoples designs as a starting point. A PD 9 cell is a good place to start. It wasn't all that steep of a trim so I wouldn't flatten it too much right off the bat. I also like flat trimmed canopies.

The suspention point will probable be farther back. I would guess .3 or more depending on the local AOA. Remember positive camber. Java should give you a good guess at the location of the CP at various AoA.

11 deg? Relitive to the zero lift line predicted by java foil? That seems a bit high to me but ok. If that's the case, given that AOA I'd use that in the rotation to create the panel set to try to set the canopy up for a constant angle of attack spanwise relative to the free stream. This means it will be a bit more like a section of a cone. The bottom skin panels will be a trapezoid, wider at the front. The top will have even more shape. I think this will let your canopy inflate more fully. If you're at that high of an AOA right now then your canopy is actually distorted slightly. It might be why you are seeing issues at diffent break settings at your end cells.

A constant AOA on a rectangular wing is not the most efficient thing in the world. The lift distribution span wise is far from a elliptical. But I would incurage you to try to keep a positive AOA all the way out to the end cells. When the canopy pitches you don't want the out side to lose lift. As I see it incorrect panel shape distorts the plane form. It kind of bends the canopy as it tryes to form it self to the surface/angle it wants to fly at. The airflow is hitting the end cell at an angle wanting to colaps it and roll the top skin under. The most extreame example would be your old "White Thing" with all rectangular panels. Remember how it just wouldn't let the top skin of the canopy open up and catch air at the end cells?

When I go back and read this these things always seem really confusing but I just don't know how to explane it any better. Or maybe I'm just full of it. Go sew some thing an see if it fucking flys. Honestly I think that's what every one else did although they will never admit to it.

Spellings going to suck on this one. If I hit preview post before I try to use their spell checker the little red underlines disappear. Ever notice that?

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
I think, at last, I'm beginning to understand what you're talking about. Generating panel shapes is not just about making the top skins wider due to the anhedral, it's actually about trying to compensate for the reduced angle of attack towards the wingtips, if I understand correctly.

I had to consider a donut shaped ram air canopy to understand this. Unless you put some twisting in the wing, it will not have an equal angle of attack over the whole wing. A normal canopy is only about a quarter of this donut but I think the same rules apply. If I have it correct, I can now see why rotating the rib about the angle of glide produces a better canopy.

As you say, this will make the bottom skins trapezoids which are wider towards the leading edge, right?

[inline rib_rotated_small2.png]

In other news I have managed to get my local library to find some copies of Poynter's manuals. It'll cost me a bit to get them out but I'll no doubt have a lot of questions answered.

Share this post


Link to post
Share on other sites
You've basically got it. The donut is a good analogy but I'm not doing a very good job of explaining what happens at the corners.

With the "donut" you can see that all the ribs are parallel to the free stream. By that I mean that the free stream is hitting them head on at their angle of attack.

If the bottom panels are rectangular then the outer ribs are not parallel to the free stream. They are actually being hit by it at an angle. A cross wind. It's actually trying to push in on and collapse the end cell. Take it to the extreme. Imagine that the bottom skin is a trapezoid with the FRONT narrower then the back by quite a bit. See how the ribs would be at an angle and the wind would be hitting the out side of the end cell.

I think you want each cell to be facing head on into the wind. You want your canopy to be a section of that "donut".

Now that's the angle relative to the free stream. The actual truth is that the local AOA at each point on the wing is different from that. That's what induced drag is. The local AOA tilting the lift vector backwards creating induced drag. That's where the lifting line/vortex lattice stuff comes in. But don't worry about that. That's just a question of the efficiency. If you want to make it fly better change the plane form to some thing more elliptical. Don't try to twist the wing. Every thing I've been talking about here is for a steady state. When you then look at how the canopy behaves when it pitches forward or back you'll find that the AOA at the center of the canopy changes more then the end cells. That's why I say that you could look at a... perturbed angle of attack as a twisted wing. As you examine that over a range of AOA you could then start to look at things like pitch stiffness of the canopy and some of the dynamics. I've never sat down and tried to do a model like that but it might be fun to try. But honestly the real answer is just to build something and jump the son of a bitch.

The spread sheet should start to make more since now. It's been a while but as I recall you can set the AOA and line length just like in your last drawing and that creates the panel shapes for you. You can enter an expected glide but that's really just about trying to generate numbers for a line set. Or best of my memory. It's kind of screwy but If you think you know your AOA that's all it really needs to draw the pattern. And when you're ready it will do this for an elliptical plane form. It kind of unrolls all the little triangles of the panel of the wing out on to a flat pattern. It's sort of a fractal integration.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
A design can only be optimized for one condition. Full flight along it's glide seems to me to be as good a choice as any.

We keep using the word twist. I'm not sure that's the right term. The line set and trim is consistent across the canopy. It's more about keeping the whole span of the wing head on into the free stream wind. You might think of it as avoiding a side slip of the end cells into the relative wind. That's an awkward way of saying it. Basically I'm advocating building the plane form so that you have a consistent angle of incidence of the wing with the local section of the wing facing head on into the local free stream flow.

Don't confuse that with the local AOA at each section which you might be able to determine with a little analysis. And when I talk about examining the wing at other angles of attack it would involve looking at the local angle of incidence when the canopy pitches. That change will not be constant along the span. When I say you might be able to treat it as a twisted wing in that case I'm saying that you might be able to use a constant spanwise twist in a wing model to approximate the change in AOI along the wing. This is not exactly correct but I'm saying that because it's a relatively solvable design that an open source peace of code could deal with. If I could find my old flight dynamics book I'd look up the one we did in class.

At this point I'd start drawing patterns and cutting fabric.

As a next step you might start looking at the aero elastic issues. For example the distortion in the wing from the inflation of the half cells. In theory if the pressure all around it were consistent the distortion in the top and bottom panels would be a function of the width and the height of the cell. Taller the cell the less arc in the skins. As you move towards the back of the canopy the rib becomes thinner, the ratio changes and the angle of the arc of the top and bottom skins increases. You could start to include this in your patterns. Note that this will trend opposite to the changes you are making right now. On the other hand ignoring this may create a certain amount of tension across the tail of the canopy and actually help to flatten the back of the cell. You might also have to look very closely at the flow around the end cell to really get an idea of the span wise pressure and tension in the canopy to give you a good idea of the distortion. That would actually be a very difficult problem to look at in detail. And I'm not sure it would give you a better canopy. Again with the just build it theory of design.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
Just when I think I understand what's going on, I'm reminded that I know nothing in the grand scheme of things. Oh well, at least I'm making an effort.

How about this then. A 'twisted' topskin with more width towards the nose, allowing the end cells to open up properly, and a rectangular bottom skin, to counter the effects of ovalisation towards the tail.

[inline parts4.png]

Share this post


Link to post
Share on other sites
I think what you're describing would be... odd. I may be just confusing the hell out of you with shit which although theoretically interesting to me is of no practical interest to you. But I'll take a shot at it. What you're basically saying there is that the change in span of the half cell, or more specifically the increase in that span for the arc of the inflation, is directly proportional to the thickness of the rib. Which is not true. In fact the rib is not a perfect wedge although the tail is close to that. The real answer would be tougher. Let's say that the out wards tension in the skin is equal to the height of the rib. What this basically says is that the pressure on the out side of the rib at the end cell is equal to the pressure on the top skin at that point. See the problem already? But bear with me. So what I'm going to say is that the angle of the top skin where it meets the rib is no longer 90 deg but "theta". I'm going to say that the ribs are still at there same places in space but stretch the top skin half cell "s" to make an arc "a" that will meet the rib at angle theta. I'm going to say that theta is a product of the forces pushing out on the top skin and out on the rib so a ratio of the half span of the cell and half span of the rib thickness. Remember there is the bottom skin doing the same on the bottom half of the rib. I'll ignore the fairly significant fact that there is less pressure differential on the bottom side for now but this really should be included. I suppose you could look at it as a ratio of the top and bottom skin pressure coefficients. Setting that aside I just want to show you how the math might work.

theta=pi/2-atan(t/2s)

"r" the radius of the arc

r=s/2sin(theta)

a=2r*theta

In truth the cell should be made from two peace's of fabric seamed together at the unloaded rib but here we are pushing all of this change to the out side edge of the cell. Then what about the trailing edge? in truth you should narrow it back down substantially to the original width. Maybe looking at the sagitta of the arc decreasing it down at the rate of a curve down to zero at the tail. Are you getting the idea that it's not a straight triangle that you're subtracting to balance the very real trapezoidal shape of the bottom skin panel? And this really isn't a fair assessment. I'm fudging hard on some of those assumptions. In truth I think you would have to look at the actual flow around the end cell, remember that there is a big ass tip vortex there, and look at the Cp along the cord top and bottom at that point. And even that is different for each cell do to the local AoA which varies from induced drag. You might also look at the way the different cells want to surge in front or lag behind as the each try to reach for there natural glide angle based on there trim to compensate for the local AOA created by the induced drag. That in it self wants to... bend the wing for or back just like the inflation wants to bend it back. Let me think... I think you'll find that it actually wants to bow the outer ends of the canopy forwards. I say that based on what I recall of the span wise lift of an untwisted rectangular wing which is far from an elliptical. So there you have something fighting the "inflation".

Is that confusing enough? That's what you get for asking me questions and making me think about shit. Personally I'd disregard all of this. As you probable should do with most of what I say. Here is my recommendation which is worth exactly as much as I'm charging you for it. IGNORE AERO ELASTIC ISSUES IN THIS ITTERATION OF YOUR DESIGN. I would build it as a rotation around the free stream line as we discussed which will give you a trapezoidal bottom skin panel and fuller top. What I think you should do on this canopy is establish you're construction techniques. What I mean by that is freeze your seams design. Go and find a real double needle sewing machine. Get a folders for your seams. How ever you're going to do them. Learn to sew and build it right. If you set this up properly it will cut the construction time down to a fraction of what it took you to build the last canopy.

You're on a good track. There is nothing wrong with you're designs. They are very conservative. They are flying fine. I would not expect major issues. I think it's time to get serious about the other half of the equation and work on being able to actually build them.

And for gods sake when I start wondering off on a tangent like these distortion issues just ignore me and go sew some thing. It's all straight out of my ass any way.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
I can't quite build an entire canopy at the moment due to a lack of that money stuff. However, I have managed to get a couple of copies of Poynter's manuals from the library, which I am certainly not scanning into my computer for later reference. I haven't read them from cover to cover yet, but there are a few things that have caught my eye.

***"In flight, the span of the ram air canopy is often shorter at the leading edge than at the trailing edge due to cell inflation"Have I got this wrong, or is this the reverse of what you were saying about aeroelastic issues, RiggerLee?

Share this post


Link to post
Share on other sites
That seem backwards to me. At least in terms of inflation. What you may be seeing is that the unloaded ribs will shift upwards. This also narrows the canopy. As the canopy slows if flare for example this will exaggerate. I'd say the distortion is a balance between inflation pressure and lift at that point of the cord which would cause more of this type of distortion at front and perticurly at low air speeds. Also if you pull down on the tail you squeeze the air out of the tail of the canopy which means the tail is able to fully expand span wise. This explains why the canopy looks so pinched, narrow at the front when you flare and land. You're out of airspeed and inflation. So At the point when you need all of the area it's actually at it's worst.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
Thank you. If you like, you can jump the canopy when it is made.

Here's a technical looking image which should hopefully make it look like I know what I'm doing, even though I have no idea.

[inline 2740_194_150_stab_new_tilted_small.png]

I think I've also decided on a name for this range of canopies, based on a similar looking older canopy. I don't think there'll be any confusion.

[inline logo_idea2.png]

In terms of where I go from here, I'm still faffing with some design details and what sort of seams I will use, but as soon as I can find somewhere to work I will begin production.

Share this post


Link to post
Share on other sites
JerryBaumchen

Are you going to have the outer lines attach to the stabilizer?

That is how it looks to me.

If so, that would put the inner A-lines farther forward that the outer A-lines during packing & in-flight.



I think you may be mistaking the chord line for the seam between rib and stabilizer.

Mark

Share this post


Link to post
Share on other sites

Yeah, as I said, it's supposed to look technical and confusing without really showing anything. ;)

On the subject of panel shape, I can't help but notice that accuracy canopies often have a trapezoidal looking planform. Is this intentional or some unwanted side effect?

Image taken from PD's facebook page.

[inline pd_cropped_sized.jpg]

Also, I finally found out what the 'STAB SLK' dimensions are on PD trim charts, and I have modified my stabiliser drawing appropriately.

[inline stab_slack.png]

Share this post


Link to post
Share on other sites
That canopy is not trapezoidal. What you're seeing is the distortion caused mainly by the unloaded ribs floating upwards when you are at very low air speed with low dynamic pressure trying to inflate the canopy. The tail is spread wide because by pulling down on it you squeeze all the air out of it, flattening it to it's actual width. Also there is more lift at the front of the canopy trying to distort it fighting against the limited inflation of the canopy trying to spread it. So span wise shrinkage is at it's worst.

Lee
Lee
[email protected]
www.velocitysportswear.com

Share this post


Link to post
Share on other sites
I have a new rib design... again. This one's for a 7 cell. I have turned to the dark side and introduced a nose lip.

[inline new_rib_small.png]

Here's where I am with the project right now. I'm saving up a bit of money so that I can buy proper materials, but one thing I don't have is anywhere to build my new canopy. Sure, I can do the sewing on my home machine in my bedroom, but I don't have anywhere to cut out the pieces. I could really do with a rigging loft of something similar.

Also, does anyone know what crossports are actually for?

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

0