this weekend I got my repairs ready and was able to go into the measurements again.
Unfortunately I was not able to recheck - but if I made no errors It will fly at least as good as before.
For now, I will give you data I already collected. The filming has to wait a little, still.
Pls read considerations recommendations carefully, I would suggest starting with the Z tweak first.
Depending on what A15 you have it may be enough to tweak Z for getting the desired performance without implementing the backweighted mixer. I unfortunately am not able to recheck because my new linelength are fixed to the new lengths of the backweighted mixer design.
This document will be updated through the next days, so table of contents is there, but is filled up by and by...
Table of contents
- The repair
- The measurements
- Considerations and Recommendations
- Fotos and Videos
Some informations to the preps:
As mentioned I had a damage with my tweaked setup. During my last ride some of the tweaking knots ripped off, causing damage to other lines (about 14 lines were gone...). I had to repair.
I decided to use Liros DC201 for the repair of all collector lines (numberlevel 2..4, egg. a56, b56, a567...). The DC 201 has a lot pull (260kg) with a very reasonable thickness (1,1mm). DC201 is the smallest line, which is 12 ply woven and has a very smooth surface, so I took these.
The Liros DC is made as a braided hose. This results in the possibility to fabric sleeves with a stichless brummel splice.
I decided to make the splice with a 20cm inner part, and the tapered section is just 6cm.
This was made, because the splice can be reopend and so you have about 10-12cm spare line to make the lines longer if they shrink over time.
The splice tool for this lines I found best to use a steel wire of 0,6mm thickness which you see at the right edge of the corner: just bend the wire.
The level one lines (number level 1 are the lines, which are connected directly to the kite (egg. a1, b3, ....) I will refabric next time I get really bored.
I would use DC 100 (100kg , 0,8mm) here.
It is a really unpleasent job to splice so many lines. But 'unfortunately' results are pretty good, because you get rid of the stepped transitions, creating more tangles and wind resistance.
BtW: Pls recheck your original lines. The wounded original lines are not well suited to do their job over time. Even my intact lines have damaged woundings - I will exchange them later on (10 lines more ....
The original lines are very touchy against torsion - so you can see the inner (white) fibres at several points at the lines, so the wounding doesnt protect them any more... pretty bad...
On this picture you see the new lines connected
On this picture you see the damaged lines: garbage!
The actual Measurements are:
B = 7,5
C = 7,5
Z = 0
I colored for the purpose of assignment:
green: the Z bridle tweak:
I cant imagine why I didn't got my tweaks a year ago werent successful. I always had kite collapses at my first tries with regular mixer setup. This tweak should work at all standard setups alone
The values make a muuuuuch better Z engagement flying at my pansh A15 in 18m. I calculated all my values based on standard mixer setting of my pansh (valued this way in delivery). Divide these values by the wingspan values to get other starting points:
For example: Z tweak of z5678 at a 9m A15:
z5678_9m = z5678_18m * Wingspan_A15_9m / Womgspan_A15_18m = -17cm * 6,65m/9,46m = -11,95cm = -12cm
Adjustments I did because it make a slight better look. This tweaks are not neccessary for stability. I did them by adding new LCLs (these little line extensions at the kite)
So forget them while starting tests
Changes to make the Backweighted Mixer run. Look: Just b78 (and connected b7 aund b8) are put to A bridle (a89, and a8, and a9).
Main point here is to add 10-12cm on the A - collector lines.
Here is a table with the measured lengths of my pansh A15/18m.
The noted deltas are the values which I found for a proper implementation of the backweighted mixer.
Pls. read the next chapter, "Considerations and recommendations".
The resleeving is made as follows:
- Old b78 is resleeved to position a89
- Old c78 is readjusted to posion b78
Here is the new line plan with the new line positions:
Pls note the resleeving:
A bridle has now the new A89 collection line (taken from old B78)
B bridle has now a new b78 line (taken from old C78)
The reason for that is shown at the tension line graphs:
Unfortunately the original lineplan is good for topology, but not good for the real position. I noted them with the Profile depth parameter (values in black).
The idea is, that the tension lines are always backward aligned in reference to direction of flight, to prevent unwanted gear effects, special in the wing tips area (note: B connection points there are NOT at 50% position as they are in center of wing, they are at 30% at the tips....)
Considerations and Recommendations
1) Start with the Z part.
The idea is to improove the kite performance by just get rid of the weak Z line tension.
Main point is to shorten z1234 by letting z5678 (mostely) its original length.
Length of z1234 should be shortened by 22,5cm.
Check stability. Perhaps you should prolong z5678 by 6cm.
Fabrication: All values shown here are based on a standard Z line (mixertest:0,0 related A): shorten z1234 and prolong z5678.
After doing this I found a week z5, which is reasonable. I shortend it (round 25% of the delta dz1234-dz5678 = -30cm, 25%*-30cm = -7cm), so you find a delta value of -7cm at my list in z5...
Why start with z?
Perhaps the performance is well enough that you are happy after tweaking just Z.
I would assume it is good enough for smaller sizes (egg A15/15m and below).
Whether it is a proper fix for A15/18m I can't check anymore, because of already implemented backweighted mixer...
(Maybe it doesnt work as good as I think while writing this, because of backweighted C role (see below)
Buts the same tweak, just start with Z.
So give me a mail, if you made a try on just tweaking Z!
General Considerations for tweaking
Welcome in the mixer hell - would be a nice title for this chapter.
Why it was so difficult for me, to make some good tweaks?
It is difficult because everything relates to each other.
One example: To shorten one line it can be achieved:
a) just by shorten it
b) just prolong all the others
Always developing, deciding/recalculating what to do is a little erroneous.
Why recalculating? When working with the original lines it is not the best idea to shorten something - you would have to knot it - so damage of the sensitive original lines would be the result.
Always rethinking things may cause errors, so take your time before moving on!
Just many lines/gears affected: Thinking about (1st and 3rd) sometimes you just 'forget' to prepare some lines in an adequate manner. Stupid point, but it hits from time to time.
Step by Step: Always give you the time to make at least a little test, after changing one thing. Condensing several tweaks to one prep is no good idea caus it may lead to total confusion (unpredicted results)
This is big: Usually all kite line considerations can be based on the mixer test setup.
This is NOT true, if you change the mixer design!
The mixer test will show you the stall position, which is fine, but you must also see the depowered position (which is even more important for stability and light wind behavior).
So usual flight conditions are nearly always NOT the mixer test position.
If you make a new gear (changing from normal mixer to backweighted) you have to rethink both, powered and depowered. For that you have to know were the lines are, when fully depowered, compared to normal mixer and backweighted.
To get into this, we have to define our desired bar throw:
So let's assume a depower stroke of 50cm.
The ratio of deviation of B in normal mixer is 25% of Z push (from powered to depowered)
A = 0 cm
B = +12,5 cm
C = +25 cm
Z = +50 cm
(based on 0,0,0,0 in mixer test condition, which is kite stall position)
The ratio of deviation of b in normal mixer is now 50% of Z pull!
A = 0 cm
B = 25 cm (<-!!!)
C = 37,5 cm
Z = 50 cm
(based on 0,0,0,0 in mixer test condition, which is kite stall position)
If you now just change the mixer from "normal" to "backweighted", you will recognize a very bad front stall collape
of kite while testing depower.
To prevent this, A Bridle has to be readjusted by 12,5 cm, so get to a proper depower position of A compared to B.
Reason: A/B is mainly responsible for kite stability in depower mode, so this should be the same as in original setup (value this is fine prepared by pansh).
We can manage this by prolonging A Bridle lines by 12(,5) cm...
C is the same, A is related to B is related to C in a proper manner like in normal mixer design, now in both, the depowered and powered position. A very surprising effekt I would say!
Indeed we see another secundary positive effect of backweighted mixer in C, because it enables better Z engagement through bar throw.... but this is beyond scope of this document for now.
Backweighted mixer - why it is better?
As mentioned the backweighted mixer is capable to handle a more open kite (power and depower it). You can see this in the possibility to handle a 12cm more open kite, being capable to go from same A/B/C depower position (like in normal mixer setup), to same stall position (nearly same mixer test)!
-> you can handle more angle of attack through your bar stroke!
Kite stability depends on profile width.
Result is: the thicker the profile is, the less support points (line connections) you need. Based on this you find the connection points on A15 not in a linear, instead of nearly progessive manner: At centercord, you find the line connection points at nearly A0% B50%, C75% Z100% ratio of profile depth.
Lock at the Picture. The Kite deviation can be linearized, if connection points match to gear ratio:
Result is, the wing profile stays more in its original shape -> by pulling the bar the air stays more in the kite which is good for shorter refresh time after (harsh) kite manoeuvres...
My recommendation for kite designers is to go for backweighted mixers in LOW AR Designs (which got a little unpopular at the moment but perhaps reincarnate when backweighted mixer is implemented in those designs.... ).
Low AR Kites are more robust and much more tolerant against line shrink missalignments - so should lead to less service requests... and you can handle those designs with the backweighted mixer better than without.
But also in Higher AR kites you can improve kite performance as long as you put the line connection points in this geometric order: You dont need the 60cm bar on big kites any more or get a better agility.
Fotos and Videos