The Dive Test

[Tim Kearsley]

I'm looking for opinions.  The dive test, intended to assist with getting the correct CG position, has been around a long time.  If I understand the principle correctly,  you trim the model for straight and level flight at a normal flying speed.  You then put the model into a dive, typically quoted as 45 degrees, and observe what happens.  If the model pulls out of the dive rapidly then your CG is likely to be too far forward, if it tucks under and increases the dive angle then the CG is too far to the rear.  The ideal is that the model maintains the dive angle or slowly pulls out. 

 

So, do you think it's a valid or indeed useful tool?  I admit that I've only ever used it personally for a couple of gliders and I've read conflicting views of its efficacy. My own view is that the importance of trimming the model properly before carrying out the test isn't emphasised enough in many descriptions.  To me, getting this right is key to the test being valid.

 

I'd value the opinions of forum users. 

 

[PatMc]

If you are referring to a power model then it's unnecessary. The point of the dive is to increase the model's airspeed above it's level cruising speed, which of course is the only way to achieve this with a glider. Obviously the same change in airspeed can be achieved with a power model simply by opening the throttle whilst flying S&L at normal cruising speed.

[Jon H]

Its total nonsense in my view for powered aircraft but understand it has some utility for gliders. 

 

Also do not forget that personal preference is a big factor. I tend to balance my warbirds 'tail heavy' by some standards and reap the reward of a model than will not spend all its time sniffing the dirt. This may mean the model is pitch sensitive at speed, so i make accommodations to deal with that. 

 

c/g is a compromise between a variety of characteristics that are good and bad. Only you can pick which positive traits are important and which negative ones you are prepared to manage as a consequence. Its the same the other way around, what bad things are you not prepared to manage and what good things are you willing to sacrifice.

 

For me, if the blasted thing cant even taxi out to the runway without tripping up and falling on its face then you are on to a looser every time. I will gladly accept a more sensitive model in the air (tamed with other settings on the radio) if it means i can take off and land without using the spinner every time. As a result i lean towards a more rearward c/g than is often considered normal. 

[Graham Bowers]

I use a variation of this test with every model I fly and find it invaluable for optimising the flight characteristics. To give you some examples of models, gangster 63, balsacraft xtreme, black horse twister / slec fun fly, Sebart MythoS, hobbyking sportjet 70. How it'd work out for models diverse from those sorts I can not say, however the aerodynamic principles still apply.

 

The variation I use is in section 4 of this  https://ita.bmfa.uk/wp-content/uploads/2023/07/Basic-Trimming-for-Aerobatics-27-07-2023-v1.pdf

 

You'll notice the document is on the BMFA ITA web page and I learned of it at an ITA event . I mention that for context as I'm attempting to improve my aerobatics so am optimising my models for that. Using this test I've shifted the CoG position back from that recommended (a bit at a time) in every case and have always found it beneficial. Better response to elevator and rudder. It's up to you how you like your model to behave in normal flight however this test gives a basis for comparison. I prefer a slight positive stability however have on one model found I liked the aerobatic responses as the CoG was adjusted rearward quite a lot , but it made the model too floaty on landing. So I moved it forward about 5mm from the rearmost test position. The point is the test gave a methodology to explore and adjust the flight characteristics to my liking.

Edited August 25, 2023 by Graham Bowers

[John Stainforth]

46 minutes ago, Tim Kearsley said:

I'm looking for opinions.  The dive test, intended to assist with getting the correct CG position, has been around a long time.  If I understand the principle correctly,  you trim the model for straight and level flight at a normal flying speed.  You then put the model into a dive, typically quoted as 45 degrees, and observe what happens.  If the model pulls out of the dive rapidly then your CG is likely to be too far forward, if it tucks under and increases the dive angle then the CG is too far to the rear.  The ideal is that the model maintains the dive angle or slowly pulls out. 

 

So, do you think it's a valid or indeed useful tool?  I admit that I've only ever used it personally for a couple of gliders and I've read conflicting views of its efficacy. My own view is that the importance of trimming the model properly before carrying out the test isn't emphasised enough in many descriptions.  To me, getting this right is key to the test being valid.

 

I'd value the opinions of forum users. 

 

I like to do tests like this, but I also check to see how much downward push I need when the plane is inverted; should only be slight. One test that I like to do, is to put the plane into a dead vertical climb to about 400 feet. Then cut the power completely and just let it fall out of the sky and take my hands off the sticks so that all the control surfaces are neutral. Very quickly the plane will be in a vertical dive. When the CG is about right I find the plane will have naturally pulled out to about 30 degrees from the vertical by the time it is about 100 feet above the ground. If it goes the other way, the CG is definitely too far back (but you probably knew that already!).

[Tim Kearsley]

Thank you gents, for your input.  Most interesting.

[Jon H]

13 hours ago, John Stainforth said:

I also check to see how much downward push I need when the plane is inverted; should only be slight

 

This is an imperfect test as at greatly depends on the wing section of the model and how it is rigged. Anything with symmetrical or semi symmetrical wing sections should be fine, but a flat bottom wing will absolutely not enjoy being flown inverted irrespective of the c/g. Even something like a Spitfire will not fly that well upside down as they were never designed to fly like that and it is prohibited to fly inverted for a long period. I never do it, but most of my warbirds need a big shove forward to keep them flying inverted should i decide to fly them that way despite them generally being balanced fairly rearward. 

 

At the end of the day there is no test that can answer the question and cover every model type. I wouldnt waste time on it. Just fly, make an adjustment and decide if it is better or worse. If better, make the same adjustment and keep going until you decide its worse now. If worse after the first adjustment go the other way and repeat the process. 

[Martin McIntosh]

Having flown in aerobatic competitions for many years it is a test which I have never bothered to try. I set the cg so that the model will just drop into a spin on full up and rudder. Works every time and gives the right sort of feel when inverted, requiring a reasonable fistful of down ele. I hate flying models with the cg so far back that elevator input is not needed because the slightest movement of the stick will make it `jumpy`. Warbirds are, of course, treated rather differently.

[Tim Kearsley]

3 hours ago, Jon - Laser Engines said:

At the end of the day there is no test that can answer the question and cover every model type. I wouldnt waste time on it. Just fly, make an adjustment and decide if it is better or worse. If better, make the same adjustment and keep going until you decide its worse now. If worse after the first adjustment go the other way and repeat the process. 

That sounds a very sensible, pragmatic approach Jon.  If you take a dozen pilots they'd probably all fly with slightly different CG positions, because it is, as you've said, a matter of preference.  My original reason for asking was that I'm in the middle of getting the CG right on a Black Magic vintage design that I built recently.  First time in the air it seemed vague and touchy to fly.  So I added a chunk of nose weight and over-egged the pudding I think, as its now sluggish and difficult to get to climb.  I'm adopting your method above now and reducing the nose weight a little at a time.

[Jon H]

Have you adjusted the rates at all? I have seen many a model wrongly laden with a ton of ballast to tame its twitchy nature. This was done as the common thought is that a twitchy model, in pitch especially, is tail heavy. In my experience this is very rarely the case and excessive rates are often the root cause of the problem. If you consider the amount of movement we need on the tx stick to make a model do something its often very tiny. If you imagine a 5p hovering on top of your stick and through a flight you never really move the stick beyond the edge of the imaginary 5p then your rates are too high. As a quick example i set my warbird aileron rates by doing a full deflection aileron roll and reducing the rates until i get to a roll rate that is a little faster than scale when the stick is hard over. I then set a higher rate above that for use in emergency's or a choppy cross wind landing when i need more control. This leaves my flying rate pretty low and gives me nice accurate control. As discussed on other threads expo is often employed excessively and incorrectly so if you have any expo on there i would wipe that out, chop the rates back and maybe even go back to the original c/g pre ballast. 

 

The black magic is a sweet little thing and should be a delight to fly. A former club mate had one and he would let me fly it back in the dim and distant past. It was most enjoyable. 

[Tim Kearsley]

Thanks Jon.  When I set up the radio for this model I adopted my usual practice and set up three rates but after actually flying it I've settled on the highest rate.  The control surfaces on the BM are fairly small.  It never did feel twitchy, rather it seemed to sit in the air with a tail-heavy "look" about it and seemed weirdly vague.  By this I mean originally we had to put in a large amount of right trim but then subsequently had to take most if not all of that out.  It seemed inconsistent.  Back on the ground, a couple of club members who are more experienced than me in flying old-timer models balanced the model on fingers and thought that the CG seemed too far back, hence the additional nose weight.  I had three flights last Thursday with the additional weight in ,and it now to me, seems sluggish and very unwilling to climb, even though there's no shortage of power.  I would add that the balance point is now forward of the point indicated on the plan.

 

I'm going to adopt your advice of removing weight a little at a time until it feels right to me.  It seems the simplest, quickest way to getting it right.

[Trevor]

I assume you have checked that the elevator is centering properly? As for the dive test, I do find it useful, although I use only a shallow dive, certainly not 45 degrees. I think it can help with a power model too. When a pilot complains that his model zooms up under power so needs more downthrust, my first question is 'are you sure the cg is right?' The dive test is one fairly direct way of checking whether the model zooms with increasing speed without any thrust line issues confusing the picture.

[Jon H]

9 hours ago, Trevor said:

I assume you have checked that the elevator is centering properly?

 

good shout, i would check the rudder as well.