Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length.

[dreadnought72]

Well done Cap - an explanation without pictures or arm-waving that I can understand!

I get you now - axis parallel to trunnions.

However ... your "*" is possibly a valid concern. If you rotate a gyro around an axis that is perpendicular to your single pivot (i.e. the hull rolling while the director's straight ahead) the gyroscope will want to precess. One way or the other. I'm never sure which.

If the effect is to pendulum-control the gun independent of roll and pitch, why not (types Andy quietly from the back) use a pendulum? Damped to avoid the crazies, heavy enough to move the pot, and swinging through the same single axis as above, while rotating with the turret? It'd save you electricity, if nothing else!

Andy,

42' launch now ready for thwarts and other internal gubbins.

[Capricorn]

Well, yes a pendulum would be much simpler, but how do you dampen it?  I thought maybe the gyro would tend to act as a dampened pendulum.  You've convinced me to consider it for sure, a handy place to hang the pendulum from would the the vertical shaft for the director turret, it's handy in that it turns as needed (the platform), and has a pretty decent vertical height if needed (see attached sketch).  I would take suggestions on how to dampen it though because if the guns teter up and down it sort of ruins the effect .

As for the gyro, I thought adding the second gimble would make it exceptionally complicated and not worth trying but looking closer it doesn't take anything extra but the second gimbal nested in the first, easy enough.  Cap 

[furball]

Well, yes a pendulum would be much simpler, but how do you dampen it?

Mount it in oil?

No idea whether this would work - just thinking of something denser than air for it to push against.


Lance

[derekwarner]

Guys....one of the issues you have here is that dreaded word SCALE 

A model of a Dreadnought [at whatever scale] with the scale 12" guns training & elevating to scale would look like .......very computer/robotic in motion...& WAY TOO FAST...something more akin to a current missile launcher or high speed 75 mm rapid fire gun

So your model vessel may pitch & roll to scale........but remember those 12" barrels in sets of three trained & elevated majestically...so this is another example where a VISUAL function cannot necessarily be scaled to provide a realstic depiction

Sorry if I appear tung tied here...another way to explain this scale anomaly is.......

We a full size humans stand on the deck of the full size vessel & view the gun train & elevate - OK

We as full sized humans view the scale model  gun train & elevate but for correctness we [the people] need to have a view & perspective to the same scale...... ...Derek

[dreadnought72]

Hi Derek - I know what you mean.

Thing is, when I go for scale speed (in order to generate a realistic wake) then the model's top speed has to be about two and a half knots. (21 divided by the square root of 72).

This means the model needs to move on the water at about 1.27 metres per second - which is 91.67 metres per second when scaled. But 21 knots is really 10.80 metres per second, so "scale time" runs 8.48 times slower. (That's the square root of 72 turning up again).

If the turrets were set to train at 34 degrees per second on the model (because it's 4 degrees a second in reality), and the hull's metacentric height was identical to the original ship, AND the model was videoed, then the video slowed down by 8.48 times - the result would be that the wake, wave speed, roll rate and turrets would all appear to move at their correct full-size speeds. That's the science bit.

But (outside films) we don't see models in slo-mo.

I agree that a watcher of the model will feel that turrets at this speed would appear to be too quick - they don't suggest a mass and inertia that we "know" was present, so I'll have to have the turrets train slower than this. The beauty of a microprocessor driving the turrets is that I can eventually set the rotation speed to any amount I like (well, up to the maximum servo speed, of course), and therefore the "art" is in finding a value which suggests "turret inertia" and to avoid turrets that look like car-assembly robots.

Ultimately, of course, this raises the feeling that the bigger the scale, the better - and (pipe dream) wouldn't it have been good if the UK hadn't trashed all our battleships? 

(Imagine how amazing it would have been to have the Warspite moored up for public visits in the 21st century...)

[Capricorn]

We know that angular velocity doesn't contain length units so it doesn't get scaled right?  The rotation rate of the model should match the actual rotation rate.  Furball, do you mean submerge the entire pendulum in oil?  That might be tricky, it would also slow the response. Cap 

[dreadnought72]

Angular velocity does contain "length units" - except these "lengths" are degrees or radians, not metres, etc...

Oil damping should work well - you want to slow the response time a little, but it could be messy and would need experimenting with. I suppose the microprocessor answer would be to feed the angle of the pendulum to some code, and "damp" short-term oscillations and oddities using software, exporting the final "preferred" pendulum angle (from some averaged angle over the last few seconds?) to a servo controlling gun angle. But that would be a little around the houses, and if done that way, then you might be better off controlling all gun angles from a single pendulum swinging in two axes, taking into account each turret's train and how that would affect it.

Complicated - but kind of neat for programmers like myself!

Regards,

Andy #1963#

[Capricorn]

Andy, angles are angles not lengths, if you multiply the radian angle by a length you get an arc length, but on the scale boat any length you multiply the angle by is already a scale length, like the length of your gun barrel.  So when the model turret turns at the actual angular velocity the scale speed of the end of the barrel moving is correct.  If you turn the model turret faster than the actual angular speed of the real turret, it is turning too fast.     However if you use the filming and slow down scenario then you would want the model turrets to turn faster.  Cap

[derekwarner]

Hi all.....&  as Andy says.........

The beauty of a microprocessor driving the turrets is that I can eventually set the rotation speed to any amount I like (well, up to the maximum servo speed, of course), and therefore the "art" is in finding a value which suggests "turret inertia" and to avoid turrets that look like car-assembly robots

 ...this will be the best scenario...where you can trail elevation & train speeds simply without great upset to any system etc.......& get the best realistic visual irrespective of what SCALE suggests it should be.....Derek 

[snowwolflair]

We know that angular velocity doesn't contain length units so it doesn't get scaled right?

Correct. rotation does not get scaled, similarly radars also run at the same rotation rate as full scale.


Think of it this way a real radar turns 400 times in a nautical mile.  So a scale radar should turn 400 times in a scale nautical mile.  You make the adjustment in the speed the ship travels so further adjustment of the rotation rate is unnecessary.  The guns are the same.

[dreadnought72]

Today's a great day.

I'm about a year into deck planking, and tonight (or certainly at some point this weekend) I will be past the halfway point. The entire port side of deck is very very very nearly complete. 

(I'm an optimist - I see this as progress!)

Zen and the Art of Deck Planking is the latest update on the front page of my site. Photos of the work should be around early next week.

Andy NA553568B

[madrob]

Well that put a smile on my face

[andrewh]

Andy NA553568B

Love your account - "diary of a lofty decker"
I had an alternative title but it seemed tasteless

Good luck with the decking - and other build tasks as the weather warms up.

What is the significance of the new handle?  are you an airfoil of 5500% thickness, or an airplane with US registration?

andrew, just andrew

[dreadnought72]

Hi Andrew,

That'd be some airfoil. ...But think of the lift!

Surely you recognise two-letters, six-digits, one-letter? A clue: I was told it before work about thirty years ago and have never forgotten it.

Odd how these numbers stick. 15 years ago I stopped working for the BBC, where - touching on The Prisoner thread - I had spent nearly a decade known as 257697B. (And at this point in proceedings, no doubt all those who did National Service will recall their numbers with absolute clarity. And yet not remember where the car keys ended up when put down fifteen minutes ago!)

As to the boat plans for the first quarter of 2009: I need to get this thing on the water. So the push will be for the decks to be done, sealed & sanded by the end of February, the hull to receive its bilge keels, and more G4 on the outside before a thorough paint job before the end of March. I need to test the electrics and ballast "as soon as", before I get moving onto the superstructure and bits.

Andy

[dreadnought72]

50%

 

[andrewh]

Andy

Well done that man!
Keep up the good work, keep us in the picture and keep us entertained with the process

OK - I don't remember mine, or perhaps I do but can't remember that I remember it or.........

We expect to be present at the launch, virtually.  Will the quaternions be sending a contingent?

andrew

[derekwarner]

As previously noted......but I must question?????? 

1) "Correct. rotation does not get scaled, similarly radars also run at the same rotation rate as full scale".

2) "Think of it this way a real radar turns 400 times in a nautical mile.  So a scale radar should turn 400 times in a scale nautical mile"

For the latter part of 2) to be correct........the leading comment in 1) must be false  ....also director rotational speed is totally independent of distance travelled by the vessel

This belongs to the same scenario as scale BLURRRR...and a radar director rotating at true scale would be just that....a BLURRRR

Think of this rotational issue in another way..... ..if propellers did not rotate in +++++ to scale RPM....the vessel would be like a near dead frog in the water....[this has no reference to the Moderator in France]

The action of a [model] propeller function is both  a mathematical & physical function to perform the desired & required scale result

The action of a [model] radar director is visual only to perform the desired & perceived result

Propellers  rotating @ 200 RPM OR 2000 RPM are a BLURRR & our human visual  perception cannot necessarily distinguish the ratio of rotational variance between both examples.....

Consider looking through the eyes of a human...if the director rotated @ 1 RPM on the real Dreadnought....try 1 RPM on the model.....Derek

[snowwolflair]

The constants are rate of angle of rotation and time, distance is scaled.

time does not get scaled, distance does, so the ship travels a shorter distance (a scale nautical mile) by sailing at scale speed.

Rate of angle of rotation in time is made up of two constants. and therefore does not change.


Another mathematician will also tell you you can scale time and leave rate of angle of rotation and distance as the constants (someone earlier talked about this by videoing in slow motion).

[Capricorn]

Andy NA553568B, Congrats on the halfway point!  I'm sure it looks great, I'd come watch the launch too but it's a bit far for me.  I like your zen piece very well, I mentioned the catlike reflexes when the xacto rolls off the table and all the response I got was to get a knife that doesn't roll (and or level my workbench).  So congrats on your ability to move quickly still too.

Sorry to belabour the angular speed bit Derek, but the rotational speed of the propellers on the model boat, as it relates to the boats speed, is a different question, definitely related to scaling but different.  I would describe the angular speed in terms of the speed that the end of the gun barrel moves when the turret turns.  If the turret and gun barrel length are all scaled properly and the angular speed of the model turret matches the real one then the linear speed of the end of the gun barrel when the turret turns would be the correct scale speed.  Whether that "appears" too slow is another question too.  Cap

[catengineman]

regarding the last few posts                   












My head hurts now!
does it matter so long as the look is what the constructor / owner wanted or even strives to get to.

R, signing off for before a mental meltdown happens

[dreadnought72]

Think of it this way a real radar turns 400 times in a nautical mile.  So a scale radar should turn 400 times in a scale nautical mile.  You make the adjustment in the speed the ship travels so further adjustment of the rotation rate is unnecessary.  The guns are the same.

Sorry for my delay in getting back to you on this - but I've been thinking. While planking. Possibly not the best combination.

Clearly we've no problem with scaling linear dimensions. All models do this. Lengths, area, volumes (and thereby masses, since we can keep densities the same) can be reduced to make a model that reflects the original vessel.

Problems arise when we want to make a working model, since we can't scale one important effect: time.

If I climbed the original Dreadnought's foretop and dropped a hammer over the edge of it, the hammer would fall 4.9 metres in one second.

On the model, I'd expect to see the (tiny) hammer fall 6.8 centimetres (=490/72) in the same period. Now the time taken for this fall under "full size" gravity is 0.118 seconds - 8.49 times quicker than I'd expect. So (the "video effect") if I ran a video of the model 8.49 times slower than "real life" I'd see what appeared to be realistic effects of gravity on the model.

8.49 is the square root of 72 (what a surprise!) and it's this time alteration which brings us to the scale speed: not surprisingly, since wave formation is tied into gravity.

We know that the scale speed is the one where the model hull generates an accurate representation of the wake of the original. For a 1/72nd scale model, where the original could steam at 21 knots, that means:

Scale speed = 21 / SQRT(72) = 2.47 knots.

The original vessel could do a nautical mile in 171 seconds.

The model's nautical mile is 1/72 times shorter, and at 2.47 knots, it takes only 20.2 seconds to cover. No surprises here, the model is 8.49 times faster.

So, to be strictly accurate, a radar (assuming the Dreadnought had one) rotating at 400 revs per 171 seconds - about 140 rpm - should be sped up to 1190 rpm.

The turrets, which could train at 4 degrees a second, should be sped up to 34 degrees per second on the model.

But clearly this is nonsense. We don't tend to view our boats via video; and whizzing radar, along with turrets moving like dingbats, doesn't read "true" to us. But that's ok, since we can't scale the viscosity of water or surface tension either - and we've all seen older films which included model boats where it's the non-scaling ability of water (big droplets, no "spray") which make us yearn for cgi to be invented.

So the answer must be: follow the "rules" for scale wakes, and apply the "art" for good looking radars and turret movement. I can live with that.

Incidentally, cgi - computer generated graphics - can be fantastic. I'd recommend watching "Ratatouille" for not only a great piece of entertainment, but for a computer generated world which rings absolutely "true". They spent, for example, an enormous amount of time to recreate the physics of liquids during the cooking sequences, and it all pays off.

(Spot the computer programming geek enjoying films.)

Best wishes,

Andy, wondering how I went from physics to rats in one post. 

[snowwolflair]

Think of it this way a real radar turns 400 times in a nautical mile.  So a scale radar should turn 400 times in a scale nautical mile.  You make the adjustment in the speed the ship travels so further adjustment of the rotation rate is unnecessary.  The guns are the same.

Sorry for my delay in getting back to you on this - but I've been thinking. While planking. Possibly not the best combination.

Clearly we've no problem with scaling linear dimensions. All models do this. Lengths, area, volumes (and thereby masses, since we can keep densities the same) can be reduced to make a model that reflects the original vessel.

Problems arise when we want to make a working model, since we can't scale one important effect: time.

If I climbed the original Dreadnought's foretop and dropped a hammer over the edge of it, the hammer would fall 4.9 metres in one second.

On the model, I'd expect to see the (tiny) hammer fall 6.8 centimetres (=490/72) in the same period. Now the time taken for this fall under "full size" gravity is 0.118 seconds - 8.49 times quicker than I'd expect. So (the "video effect") if I ran a video of the model 8.49 times slower than "real life" I'd see what appeared to be realistic effects of gravity on the model.

8.49 is the square root of 72 (what a surprise!) and it's this time alteration which brings us to the scale speed: not surprisingly, since wave formation is tied into gravity.

We know that the scale speed is the one where the model hull generates an accurate representation of the wake of the original. For a 1/72nd scale model, where the original could steam at 21 knots, that means:

Scale speed = 21 / SQRT(72) = 2.47 knots.

The original vessel could do a nautical mile in 171 seconds.

The model's nautical mile is 1/72 times shorter, and at 2.47 knots, it takes only 20.2 seconds to cover. No surprises here, the model is 8.49 times faster.

So, to be strictly accurate, a radar (assuming the Dreadnought had one) rotating at 400 revs per 171 seconds - about 140 rpm - should be sped up to 1190 rpm.

The turrets, which could train at 4 degrees a second, should be sped up to 34 degrees per second on the model.

But clearly this is nonsense. We don't tend to view our boats via video; and whizzing radar, along with turrets moving like dingbats, doesn't read "true" to us. But that's ok, since we can't scale the viscosity of water or surface tension either - and we've all seen older films which included model boats where it's the non-scaling ability of water (big droplets, no "spray") which make us yearn for cgi to be invented.

So the answer must be: follow the "rules" for scale wakes, and apply the "art" for good looking radars and turret movement. I can live with that.

Incidentally, cgi - computer generated graphics - can be fantastic. I'd recommend watching "Ratatouille" for not only a great piece of entertainment, but for a computer generated world which rings absolutely "true". They spent, for example, an enormous amount of time to recreate the physics of liquids during the cooking sequences, and it all pays off.

(Spot the computer programming geek enjoying films.)

Best wishes,

Andy, wondering how I went from physics to rats in one post. 

Yes but you are mixing apples with pears.  Gravity and viscosity may have bearing on scale but have nothing to do with roataional velocities.

I if you expand your mixing of constants and variables to its logical conclusion, you should include the lunar tides, the Gregorian calender, not to mention the slowing of the roataion of the earth, and the colour of the ships cat. 

First rule of engineering   KISS - Keep It Simple Stu***.

[dreadnought72]

A brief update to the ship's boats has been added.

Deck planking, meanwhile, is progressing at what can only be described as alarming speed. I'm well past halfway, moving outwards and forwards along the starboard side. X and Y turrets are engulfed in a sea of woodwork, the rear of the superstructure cut-out is about to follow.

Hmmm...a quick dash past Q turret, the forecastle's starboard side, and I can almost imagine the end!

....In a week or three. 

Andy

[dreadnought72]

And one week on I'm at:

...80%, at least.

New pictures up tomorrow. Not that there's much new to see really, other than planks. And caulking. And, errr, that's it.

But that will give me the opportunity to mention an RSI from PVA-bottle lifting, along with what could be a dose of Dupuytren's contracture.

  (...which is very possibly the first thing I've ever had in common with Margaret Thatcher.)

Andy, finishing plank in sight!

[Capricorn]

Andy, Good to hear you are moving along, interesting how these things go in spurts, I'm in the dumps now, not getting anywhere.  I used CA when I was planking my DD hull (not near as many planks) and the fumes were most unpleasant, so at least you've avoided that.  Cap