DLG flying skills

[tomc]

So....

3 grams of weight in the nose brings forward the C of G by around 3mm, which makes a big difference. In flat calm conditions it now continues flying forwards at a more normal speed (still seems fairly bullet like), and not every control input making it stall, even managed to catch it twice (out of about 20 goes). I'm surprised at just how sensitive it seems to be to such tiny adjustments, am tempted to put a few more grams in and see if that improves it more.

Thanks

Tom

[Paul Gleeson]

go bit by bit, gram by gram

[StraightEdge]

Well done for making a start on that!

A bit more methodical detail now...

Working just in speed-mode for the time being, get as much height as possible on launch, then pitch forward into level flight with normal speed on (push over before the very top of the climb-phase else you'll slow too much and stall, losing a lot of height at the outset), then:

1. What does it do in this level flight with the new current CG?  If it clearly slows down and wants to stall, then add some down-trim (do this is stages over successive launches) until it stops any stall-like behaviour.  If it noses down and speeds up too much then add up-trim in stages.
2. Now, with the speed-mode level-flight trim temporarily sorted out and left alone, what does it then do if immediately after pitching forward into level flight you then put it into shallow dive (30°) and centre the three main controls?  Continues on a straight trajectory, pulls up , or pulls down?  If it is pulling up, then your CG is too forward (i.e. you've been needing up-elevator trim to keep it level in normal flight), if pulling down then CG isn't yet forward enough (needing down-trim for level flight). 
3. Add or remove nose-weight as necessary but in small steps, each time repeating Step (1) level flight elevator-trims, then Step (2) dive-tests, until you've got a nice pace in level flight but not too fast, and just a very, very slight pull-up on the shallow dive-test.

Once you've got the CG in the right ball park in speed-mode, leave the CG alone and move on to trimming the elevator in level flight only for each of cruise- and float-modes.   Because of the additional camber in these modes, there's an increased risk of stalling by over-controlling and/or when high up, so it might be an idea - once you've got the trim for each just right - to add a click or two of down-trim as an insurance policy.

Note:  make sure your wing-camber for each of these two slower modes (as well as speed mode of course ) is exactly as per the instructions for your model, but it is perhaps better to err on the side of a smidge too little for each than too much at this stage, and make sure the down-travel for each camber setting is exactly the same for both wings - else this will make trimming the elevator for each mode much harder.

Also experiment with the camber/reflex and elevator settings for launching, especially zoom-mode (the climb).  To keep things simple, you can have this the same as speed-mode at first, but tweaking it might get you launching higher and buy you more time for the glide tests.... then positioning the model for landings/catches.

The whole process of sorting out the trim of any precision model is a very methodical process:  make only one change at a time and repeat each of the tests in sequence until you're as close as possible to the ideal trim.

If possible do all this in 'dead-air', i.e. in calm weather, early on in the day before any thermal activity (both lift and sink) starts to build, or much later when everything has calmed down.

[mikef]

9 hours ago, StraightEdge said:

........

Now, with the speed-mode level-flight trim temporarily sorted out and left alone, what does it then do if immediately after pitching forward into level flight you then put it into shallow dive (30°) and centre the three main controls?  Continues on a straight trajectory, pulls up , or pulls down?  If it is pulling up, then your CG is too forward (i.e. you've been needing up-elevator trim to keep it level in normal flight), if pulling down then CG isn't yet forward enough (needing down-trim for level flight). 

......

You want it to be pitching up a little from the downward path in the dive test just not too much.  'How much' is a matter of taste.  I like to see a steady glide for a few seconds before I start a dive test.
l don't understand your statement, “you've been needing up-elevator trim to .....”,     To get an initial smooth glide with a new model, it might need up, down or no change in elevator position.

   

[StraightEdge]

42 minutes ago, mikef said:

l don't understand, “you've been needing up-elevator trim to a keep it level”...

I said this in the context of if the model's CG was too forward, then said that the opposite, i.e. down-elevator trim, would have been needed in level flight if the CG was too far back.

I was really just repeating the key idea that had been covered in previous posts:  inserting it again as a reminder in the more detailed sequence of steps I was trying to set out, my main point being that is all an iterative process composed of discrete, repeated steps.  I could have maybe expressed the whole thing better - more concisely perhaps -  but it was getting late into the night!

 

[StraightEdge]

Tom, and any other newbies reading this...

Go here for chapter and verse from Thomas at Armsoar on his methodical sequence of setting up a DLG:

Armsoar DLG Setup Guide

 

[mikef]

I  want to be sure that nobody gets the wrong idea here - 'up elevator trim' and 'down elevator trim' are not statements about where the elevator is relative to ‘lined up with the tailplane‘ - that doesn't matter much.   A model needs 'up-elevator trim' if you need to slow it down (and vice versa).

If I take a new model to the field and I have to give it two clicks of down trim to stop it stalling and get a nice glide, it does not mean the CG is too far aft.   (Or vice versa)

It's quite possible that I might have needed to give that same new model two clicks of 'up' to get the same good glide.  It depends on where I set neutral on the elevator.
It will not generally be the case that the elevator lines up exactly with the tailplane in any trimmed state and, if you have various wing cambers in different flight phases, it will not be lined up exactly in all of them.

(The so called 'trim drag'  this deflection causes is negligible but you can get the elevator to zero in at least one chosen phase  by shimming the fixed part of the tailplane.  Personally, I think the time is better spent practising your flying.)

[MikeDLG]

I just want to mention that for me at least the dive test is not the way to decide upon my CG position.  For me, I adjust the CG until I get the model flying in a way I’m happy with.    I use the dive test but I wouldn’t simply add or remove nose weight purely based on dive test behaviour.  I tend to like a more stable glider than the average, I think, so my gliders tend pull out of the dive noticeably.  
 

For the less experienced, it is a useful tool to set a good ball park CG position  provided it is done correctly.  The most important thing to do is to trim for level flight before starting the dive. Results are meaningless otherwise. Of course, level flight with a glider is a gentle decent, just to complicate matters

[David North]

Hello, I’ve just started to teach myself to fly an Elf DLG. I have some technical knowledge but zero model flying experience, so I’m trying to trim the model and learn to fly it at the same time.

From a gentle side launch (no foot movement) my Elf tends to pitch up and stall or loop into the ground, although it glides nicely and the cg position doesn’t seem excessively forward at 73mm behind the leading edge apex - 74/75 mm is recommended. I can control this behaviour by reaching across with my left hand and holding a significant amount of down elevator during launch, but this doesn’t seem right. Is it possible that poor launch technique is the problem? Or is it more likely due to excess stability/decalage?

[mikef]

We use a two stage trim for the launch (usually implemented by momentary buttons/switches)

The second,  'straight up', part of the launch is a trim that has the model flying almost vertically without pitching up or down from a flat climb - this will have 'down elevator' compared to the glide.  The first, ‘pitch up' phase converts the near horizontal realease to a climb at your chosen angle - this will have up elevator relative to the glide.  This up elevator is released at or soon after launch - adjusted to get the pitch-up you want.

If the model is looping into the ground, it sounds as if you could move the CG back a couple of millimetres.  My Elf CG is at 74mm behind the wing leading edge apex but with a two stage climb trim you set the CG to give the pitch stability that suits you in the glide and trim both phases of the launch independently.

Can you organise your Tx to produce the two-stage climb trim and then go to glide trim after the bunt?

 

[David North]

Thanks Mike. I’ll see if I can cope with the cg back a bit. It surprises me that some people claim to fly their Elves with the cg much further forward - I wonder how they avoid “my” problem?

I’ll move the momentary switch on my Taranis to the left, so that I can reach it with my left hand during my right-handed launch. I need to practise soldering before I attempt this!

if I can get my launches to last a bit longer I’ll hopefully have time to get my right hand onto the transmitter so that I can bunt it at the top of the climb and/or select a different flight mode.
 

Many thanks

David

[StraightEdge]

David, I swapped the the spring-loaded switch on my Taranis from the right (SH) to the left (SF), which did involve a bit of careful de- and re-soldering using a 25W iron, but not hard.  (The current X9DPlus, from 2019 onwards, now has an additional momentary push-button on the left, but it is inboard of the corner which is harder to reach than if it was properly on the edge!)

A short-term solution (while you also practice putting a bit more twist into your back for more speed on release) could be to use the existing left-hand switch flicked into launch-mode (i.e pitch-up) before you swing, with your finger ready to flick it forward as soon as you release the Elf.  Cumbersome but not half as impossible as trying to get your right-hand back to the transmitter!

Also, where is your hand pointing after release - at the horizon or up?  My launches became much higher when I stopped trying to launch my Elf up.

[mikef]

David, your Taranis can do a lot so it should be easy to get the two stage climb trim.  Lots of pilots on here do it.  I only speak Multiplex so I'm not much use.

[David North]

2 hours ago, StraightEdge said:

David, I swapped the the spring-loaded switch on my Taranis from the right (SH) to the left (SF), which did involve a bit of careful de- and re-soldering using a 25W iron, but not hard.  (The current X9DPlus, from 2019 onwards, now has an additional momentary push-button on the left, but it is inboard of the corner which is harder to reach than if it was properly on the edge!)

A short-term solution (while you also practice putting a bit more twist into your back for more speed on release) could be to use the existing left-hand switch flicked into launch-mode (i.e pitch-up) before you swing, with your finger ready to flick it forward as soon as you release the Elf.  Cumbersome but not half as impossible as trying to get your right-hand back to the transmitter!

Also, where is your hand pointing after release - at the horizon or up?  My launches became much higher when I stopped trying to launch my Elf up.

Yes, that’s what I intend with the SH and SF switches. A shame that it requires soldering.

The strange thing is that I need down elevator on launch, not up elevator. I suspect that I may releasing the model a bit tail-down, although it isn’t deliberate. I intend to try a more committed swing. Do people generally do a full discus launch with the Elf?

[mikef]

32 minutes ago, David North said:

Yes, that’s what I intend with the SH and SF switches. A shame that it requires soldering.

The strange thing is that I need down elevator on launch, not up elevator. I suspect that I may releasing the model a bit tail-down, although it isn’t deliberate. I intend to try a more committed swing. Do people generally do a full discus launch with the Elf?

Down elevator will stop it looping and it is what you will have for the steep bit of straight climb.  When you get that straight bit sorted, you need a bit of up elevator, relative to 'straight climb',  for a moment on release to pitch up, from your slightly above horizontal release, to the steep bit.  Glide trim comes in after you bunt.

Don't try to alter the pitch with  your launching hand- you will damage the wing tip.  You do use a full-out launch with any DLG if you want maximum height.  Aim at a point on the horizon when you launch - the model will naturally fly up slightly as you release it

The following are the tailplane trailing edge deflections for my Elf.  Zero is the glide setting.  It!s an all-moving tail so deflections are small.   74mm cg.

Pitch-up on release, held until the moment of release...     0.5 mm down

Thus then moves to steep climb no pitching up or down...     1mm down.

Bunt then  glide at....   0 mm

So, and I had forgotten, my Elf does pitch up even with 0.5 mm down elevator relative to the glide.  In the glide setting it will pitch up too strongly - your current situation.  

My conclusion here is that I might want to move the cg back from the current 74mm behind LE apex.  Having said that, my Elf has a 1hour flight on it in this trim so it's quite flyable!

 

[David North]

1 hour ago, mikef said:Pitch-up on release, held until the moment of release...     0.5 mm down

Mike, did you mean “0.5 mm up”? Anyway, thanks for this data, really gives me something to go on.

 

[mikef]

No, 0.5 mm down.  If I put the tail TE 0.5 mm up, my Elf would Do a tight loop off the launch for sure.

This isn't what I expected to find.  Here are the elevator deflections for a Snipe I've been flying for years.  I measured them the other day to help me set up a new one.  Relative to glide,  pitch-up after launch 1mm up.  Steep flat climb, 1mm down.

The bottom line is that the up steep and flat setting will have down elevator relative to glide.  The pitch-up setting will have up elevator relative  to the up steep and flat setting.  You will adjust these settings on your particular model to get the effect you want.

 

 

[David North]

1 hour ago, mikef said:

No, 0.5 mm down.  If I put the tail TE 0.5 mm up, my Elf would Do a tight loop off the launch for sure.

This isn't what I expected to find.  Here are the elevator deflections for a Snipe I've been flying for years.  I measured them the other day to help me set up a new one.  Relative to glide,  pitch-up after launch 1mm up.  Steep flat climb, 1mm down.

The bottom line is that the up steep and flat setting will have down elevator relative to glide.  The pitch-up setting will have up elevator relative  to the up steep and flat setting.  You will adjust these settings on your particular model to get the effect you want.

 

 

Hopefully I’m starting to get my head around this. So am I right in thinking that your Snipe probably has less stability (as in dive test) than your Elf, as currently configured? Maybe a low-inertia, lightweight model like the Elf is just too twitchy with the cg further aft?

[David North]

I’ve now swapped the non-latching switch to the left of my tx and followed mikef’s suggestions regarding elevator offsets for the various flight modes and cg position.

Now waiting for calm weather to give it a go.