I want to "hit" a fixed spot with a physical object over
relatively short distances (< ~20 ft).
This must not present a danger to nearby bystanders (in the
event of a "misfire"). "Weapons" are out of the question.
And, the object must be of sufficient size to be clearly
visible in transit. This also suggests a low transit
velocity.
I figure I need a bit of mass to ensure aerodynamic
effects don't bugger the calculations. E.g., a softball
would be better than a softball-sized hollow ball
which might exhibit more nonlinear behaviors as it
transits from projectile to ballistic motion.
The target is (effectively) a "spot on the floor".
I.e., not a vertical "hoop" to pass through (like
goalposts in soccer).
Accelerating a significant mass would likely prove to
be a challenge so "lobbing" the object seems more
practical. It would also *seem* to be more tolerant
of aiming issues than something HOPING to travel in a
straight line (like a bullet).
So, conceptually, a tube ("barrel") to guide the
initial segment of flight and some sort of mechanism
to propel the object from the tube.
Pneumatic, hydraulic, mechanical, etc.
Will the control over the propulsive force be the tougher
challenge or the precise aiming of the launch tube?
[You are given (r,theta) to target and no feedback as
to proximity of strike -- unless a direct strike. The
target -- or launcher -- will move after each attempt]
Of course, "you" is a machine...
Short range targeting
Started by ●December 25, 2021
Reply by ●December 25, 20212021-12-25
On Saturday, December 25, 2021 at 6:47:51 PM UTC-6, Don Y wrote:> I want to "hit" a fixed spot with a physical object over > relatively short distances (< ~20 ft). > > This must not present a danger to nearby bystanders (in the > event of a "misfire"). "Weapons" are out of the question. > > And, the object must be of sufficient size to be clearly > visible in transit. This also suggests a low transit > velocity. > > I figure I need a bit of mass to ensure aerodynamic > effects don't bugger the calculations. E.g., a softball > would be better than a softball-sized hollow ball > which might exhibit more nonlinear behaviors as it > transits from projectile to ballistic motion. > > The target is (effectively) a "spot on the floor". > I.e., not a vertical "hoop" to pass through (like > goalposts in soccer). > > Accelerating a significant mass would likely prove to > be a challenge so "lobbing" the object seems more > practical. It would also *seem* to be more tolerant > of aiming issues than something HOPING to travel in a > straight line (like a bullet). > > So, conceptually, a tube ("barrel") to guide the > initial segment of flight and some sort of mechanism > to propel the object from the tube. > > Pneumatic, hydraulic, mechanical, etc. > > Will the control over the propulsive force be the tougher > challenge or the precise aiming of the launch tube? > > [You are given (r,theta) to target and no feedback as > to proximity of strike -- unless a direct strike. The > target -- or launcher -- will move after each attempt] > > Of course, "you" is a machine...A horse shoe probably wouldn't be what you're after. A bean bag for a corn hole game? Build a gun using electrical conduit? An 1 1/4" piece with end caps on it might fit tightly enough in a 2" or 2 1/2" to be blown out with an air gun.
Reply by ●December 26, 20212021-12-26
Don Y <blockedofcourse@foo.invalid> wrote:> I want to "hit" a fixed spot with a physical object over > relatively short distances (< ~20 ft). > > This must not present a danger to nearby bystanders (in the > event of a "misfire"). "Weapons" are out of the question. > > And, the object must be of sufficient size to be clearly > visible in transit. This also suggests a low transit > velocity. > > I figure I need a bit of mass to ensure aerodynamic > effects don't bugger the calculations. E.g., a softball > would be better than a softball-sized hollow ball > which might exhibit more nonlinear behaviors as it > transits from projectile to ballistic motion. > > The target is (effectively) a "spot on the floor". > I.e., not a vertical "hoop" to pass through (like > goalposts in soccer). > > Accelerating a significant mass would likely prove to > be a challenge so "lobbing" the object seems more > practical. It would also *seem* to be more tolerant > of aiming issues than something HOPING to travel in a > straight line (like a bullet). > > So, conceptually, a tube ("barrel") to guide the > initial segment of flight and some sort of mechanism > to propel the object from the tube. > > Pneumatic, hydraulic, mechanical, etc. > > Will the control over the propulsive force be the tougher > challenge or the precise aiming of the launch tube? > > [You are given (r,theta) to target and no feedback as > to proximity of strike -- unless a direct strike. The > target -- or launcher -- will move after each attempt] > > Of course, "you" is a machine...I'm loving the slip into alcoholism posts you make.
Reply by ●December 26, 20212021-12-26
On Saturday, December 25, 2021 at 10:01:03 PM UTC-5, Jasen Betts wrote:> On 2021-12-26, Don Y <blocked...@foo.invalid> wrote: > > I want to "hit" a fixed spot with a physical object over > > relatively short distances (< ~20 ft). > > Accelerating a significant mass would likely prove to > > be a challenge so "lobbing" the object seems more > > practical. It would also *seem* to be more tolerant > > of aiming issues than something HOPING to travel in a > > straight line (like a bullet). > > > > So, conceptually, a tube ("barrel") to guide the > > initial segment of flight and some sort of mechanism > > to propel the object from the tube. > mini trebuchet+ 1 -- Rick C. - Get 1,000 miles of free Supercharging - Tesla referral code - https://ts.la/richard11209
Reply by ●December 26, 20212021-12-26
On 12/25/2021 6:55 PM, Dean Hoffman wrote:> On Saturday, December 25, 2021 at 6:47:51 PM UTC-6, Don Y wrote: >> I want to "hit" a fixed spot with a physical object over >> relatively short distances (< ~20 ft). >> >> This must not present a danger to nearby bystanders (in the >> event of a "misfire"). "Weapons" are out of the question. >> >> And, the object must be of sufficient size to be clearly >> visible in transit. This also suggests a low transit >> velocity. >> >> I figure I need a bit of mass to ensure aerodynamic >> effects don't bugger the calculations. E.g., a softball >> would be better than a softball-sized hollow ball >> which might exhibit more nonlinear behaviors as it >> transits from projectile to ballistic motion. >> >> The target is (effectively) a "spot on the floor". >> I.e., not a vertical "hoop" to pass through (like >> goalposts in soccer). >> >> Accelerating a significant mass would likely prove to >> be a challenge so "lobbing" the object seems more >> practical. It would also *seem* to be more tolerant >> of aiming issues than something HOPING to travel in a >> straight line (like a bullet). >> >> So, conceptually, a tube ("barrel") to guide the >> initial segment of flight and some sort of mechanism >> to propel the object from the tube. >> >> Pneumatic, hydraulic, mechanical, etc. >> >> Will the control over the propulsive force be the tougher >> challenge or the precise aiming of the launch tube? >> >> [You are given (r,theta) to target and no feedback as >> to proximity of strike -- unless a direct strike. The >> target -- or launcher -- will move after each attempt] >> >> Of course, "you" is a machine... > > A horse shoe probably wouldn't be what you're after. > A bean bag for a corn hole game?A bean bag has several useful qualities as a projectile: - weight, size and density are all adjustable - shape can be tailored (with/without dynamic consequences) - "skin" can be adjusted to facilitate travel through launcher - skin can support different launching technologies (e.g., pneumatic/hydraulic without intervening "piston") - easy to "dampen" its response on hitting another object (including a "miss") as well as audibly "indicating" - can be made highly visible, even with diminished size - *likely* well tolerated if accidentally striking a bystander Reloading may be a bit of a challenge. But, I suspect you could design a magazine that held some number of them and facilitated *reliable* loading of successor "rounds" as the launcher/target are being moved. OTOH, the damped response means misses will tend to pile up. This might require some other mechanism to clear them away.> Build a gun using electrical conduit? An 1 1/4" piece with end caps on it > might fit tightly enough in a 2" or 2 1/2" to be blown out > with an air gun.I think that would be hazardous to bystanders. Imagine the machine's targeting malfunctioning and mistakenly hitting a bystander. A slug of *water* is almost ideal in that it "disappears" after it makes contact (hit *or* miss). And, range could be adjusted directly via (water) pressure. But, pretty messy. It would also require a higher launch velocity to remain intact in transit (a stream wouldn't be acceptable) And, would likely be difficult to make visible (unless you relied entirely on the final "splash")
Reply by ●December 26, 20212021-12-26
On 12/25/2021 7:45 PM, Jasen Betts wrote:> On 2021-12-26, Don Y <blockedofcourse@foo.invalid> wrote: >> I want to "hit" a fixed spot with a physical object over >> relatively short distances (< ~20 ft). > >> Accelerating a significant mass would likely prove to >> be a challenge so "lobbing" the object seems more >> practical. It would also *seem* to be more tolerant >> of aiming issues than something HOPING to travel in a >> straight line (like a bullet). >> >> So, conceptually, a tube ("barrel") to guide the >> initial segment of flight and some sort of mechanism >> to propel the object from the tube. > > mini trebuchet, catapult, or springboard.I thought of this with bean bags but think it is more complicated than other approaches (because you don't need to be able to scale up) I'm not sure you (a machine) can reliably adjust the sling's release point (or spring tension) to achieve a particular striking distance. Without feedback, I think you need a mechanism that is highly (open-loop) predictable as it (the machine) can't learn from previous attempts (because the target will have moved in range and azimuth). I think getting the azimuth correct is relatively easy -- assuming the centerline of the "barrel" truly represents the initial arc segment (how likely for a trebucher to suffer a bias to one side or the other, depending on how the sling release is designed?) And, I'm not sure how the machine would reload itself though I imagine a mechanism is possible due to the reduced scale. There'd also have to be some mechanical stops to prevent excessive projectile speed and/or overthrow.>> Pneumatic, hydraulic, mechanical, etc. >> >> Will the control over the propulsive force be the tougher >> challenge or the precise aiming of the launch tube? > > friction on the insides of the tube unless you use plenty of lube or > some sorte of sabot. (hence my favouring open designs)That would depend on the nature of the projectile and it's outer surface (as well as the inner surface of such a "barrel"). A beanbag made out of a "fluffy" material would likely exhibit a predictable drag -- and the length of the barrel doesn't change. Something ejected with a liquid "expellant" would rely on the liquid's consistency between shots to make the transit predictable. [If the barrel is short enough, any "waste expellant" could possibly be captured directly in front of the barrel's terminus.] The predictability aspect IMO is the driving issue. Hence the question as to where the "sensitivity" in the design lies. Even if you design the mechanism and projectiles to be repeatable, you still have no way of knowing how ANY of them performed, in the wild. So, any "calibration" has to happen at time of design/manufacture and hold true thereafter. [You can observe the process *until* the projectile exits the mechanism but not thereafter. E.g., you could "watch" the trebuchet's arm's motion and release point to verify it is happening as intended. But, of course, that defines an historic event that can no longer be influenced!] Note that this can be a complex algorithm as a machine is performing the calculation(s) without the need for an operator. But, the calculation has to *remain* valid, thereafter, as there's no operator to tweek it based on observations! [Of course, there may be some value to a randomizing strategy in that it may "get lucky" more often than a repeatable one that is "always wrong"!]
Reply by ●December 26, 20212021-12-26
On 26/12/2021 07:30, Rick C wrote:> On Saturday, December 25, 2021 at 10:01:03 PM UTC-5, Jasen Betts wrote: >> On 2021-12-26, Don Y <blocked...@foo.invalid> wrote: >>> I want to "hit" a fixed spot with a physical object over >>> relatively short distances (< ~20 ft). >>> Accelerating a significant mass would likely prove to >>> be a challenge so "lobbing" the object seems more >>> practical. It would also *seem* to be more tolerant >>> of aiming issues than something HOPING to travel in a >>> straight line (like a bullet). >>> >>> So, conceptually, a tube ("barrel") to guide the >>> initial segment of flight and some sort of mechanism >>> to propel the object from the tube. >> mini trebuchet > > + 1Yes. They are great fun. There is a bloke near me who has scale model fully operational trebuchet able to lob a water melon about 200m! It is very impressive and apart from the beam quite compact. When things return to normal I might try and catch up with him again to film it in action. He does charity gigs with it from time to time in normal summers - needs a *lot* of space. (and careful crowd control down range) This is the biggest one I have seen online and representative of the sort of design that would work. Camera work is pretty good too: https://www.youtube.com/watch?v=UdGqggET0o4 -- Regards, Martin Brown
Reply by ●December 26, 20212021-12-26
On 2021-12-26, Don Y <blockedofcourse@foo.invalid> wrote:> On 12/25/2021 7:45 PM, Jasen Betts wrote: >> On 2021-12-26, Don Y <blockedofcourse@foo.invalid> wrote: >>> I want to "hit" a fixed spot with a physical object over >>> relatively short distances (< ~20 ft). >> >>> Accelerating a significant mass would likely prove to >>> be a challenge so "lobbing" the object seems more >>> practical. It would also *seem* to be more tolerant >>> of aiming issues than something HOPING to travel in a >>> straight line (like a bullet). >>> >>> So, conceptually, a tube ("barrel") to guide the >>> initial segment of flight and some sort of mechanism >>> to propel the object from the tube. >> >> mini trebuchet, catapult, or springboard. > > I thought of this with bean bags but think it is more complicated > than other approaches (because you don't need to be able to scale > up) > > I'm not sure you (a machine) can reliably adjust the sling's release > point (or spring tension) to achieve a particular striking distance. > > Without feedback, I think you need a mechanism that is highly > (open-loop) predictable as it (the machine) can't learn from previous > attempts (because the target will have moved in range and > azimuth). > > I think getting the azimuth correct is relatively easy -- assuming > the centerline of the "barrel" truly represents the initial arc > segment (how likely for a trebucher to suffer a bias to one side > or the other, depending on how the sling release is designed?) > > And, I'm not sure how the machine would reload itself though > I imagine a mechanism is possible due to the reduced scale. > > There'd also have to be some mechanical stops to prevent > excessive projectile speed and/or overthrow. > >>> Pneumatic, hydraulic, mechanical, etc. >>> >>> Will the control over the propulsive force be the tougher >>> challenge or the precise aiming of the launch tube? >> >> friction on the insides of the tube unless you use plenty of lube or >> some sorte of sabot. (hence my favouring open designs) > > That would depend on the nature of the projectile and it's > outer surface (as well as the inner surface of such a "barrel"). > A beanbag made out of a "fluffy" material would likely exhibit > a predictable drag -- and the length of the barrel doesn't change. > > Something ejected with a liquid "expellant" would rely on the > liquid's consistency between shots to make the transit predictable. > [If the barrel is short enough, any "waste expellant" could > possibly be captured directly in front of the barrel's terminus.] > > The predictability aspect IMO is the driving issue. Hence the > question as to where the "sensitivity" in the design lies. > > Even if you design the mechanism and projectiles to be repeatable, > you still have no way of knowing how ANY of them performed, in the > wild. So, any "calibration" has to happen at time of design/manufacture > and hold true thereafter. > > [You can observe the process *until* the projectile exits the > mechanism but not thereafter. E.g., you could "watch" the > trebuchet's arm's motion and release point to verify it > is happening as intended. But, of course, that defines an > historic event that can no longer be influenced!]You can measure the speed and angle of the arm when the it releases the projectile from that you can make a fairly good guess at its trajectory.> Note that this can be a complex algorithm as a machine is > performing the calculation(s) without the need for an operator. > But, the calculation has to *remain* valid, thereafter, as there's > no operator to tweek it based on observations! > > [Of course, there may be some value to a randomizing strategy > in that it may "get lucky" more often than a repeatable one > that is "always wrong"!]Yeah if the target is not found it becomes a blind hunt like a game of battleshipa. But so long as the real game is not a type of crazy golf it's just a matter of picking the right search algorithm. -- Jasen.
Reply by ●December 26, 20212021-12-26
On 26/12/2021 08:17, Don Y wrote: <snip>> > A slug of *water* is almost ideal in that it "disappears" after it > makes contact (hit *or* miss). And, range could be adjusted directly > via (water) pressure. > > But, pretty messy. It would also require a higher launch velocity to > remain intact in transit (a stream wouldn't be acceptable) > > And, would likely be difficult to make visible (unless you relied > entirely on the final "splash")Snowballs. But messy. Is there something with similar mechanical properties which doesn't melt? That way you could form it into different sizes as needed. -- Cheers Clive
Reply by ●December 26, 20212021-12-26
On 12/26/2021 4:07 AM, Jasen Betts wrote:> On 2021-12-26, Don Y <blockedofcourse@foo.invalid> wrote:>> The predictability aspect IMO is the driving issue. Hence the >> question as to where the "sensitivity" in the design lies. >> >> Even if you design the mechanism and projectiles to be repeatable, >> you still have no way of knowing how ANY of them performed, in the >> wild. So, any "calibration" has to happen at time of design/manufacture >> and hold true thereafter. >> >> [You can observe the process *until* the projectile exits the >> mechanism but not thereafter. E.g., you could "watch" the >> trebuchet's arm's motion and release point to verify it >> is happening as intended. But, of course, that defines an >> historic event that can no longer be influenced!] > > You can measure the speed and angle of the arm when the it releases > the projectile from that you can make a fairly good guess at its > trajectory.Yes, but you would have wanted to be controlling that at the same time; so, measurement just verifies your control performed as intended.>> Note that this can be a complex algorithm as a machine is >> performing the calculation(s) without the need for an operator. >> But, the calculation has to *remain* valid, thereafter, as there's >> no operator to tweek it based on observations! >> >> [Of course, there may be some value to a randomizing strategy >> in that it may "get lucky" more often than a repeatable one >> that is "always wrong"!] > > Yeah if the target is not found it becomes a blind hunt like a game of > battleshipa.Except the targets don't move in Battleship. So, you could make several attempts to hit a specific physical location by tweeking your control, even if it is inaccurate. I.e., "I want to hit a spot 12 ft from me in this direction. Set control to 12, fire. If not hit, set control to 13, fire. If not hit, set control to 11, fire. Hopefully, one of those three control settings actually produces a 12 ft throw!" If "12" doesn't always yield the same throw, then all bets are off. And, if the target can be 18 ft away when you take your next shot, then the control may as well be random as the effect is unpredictable.> But so long as the real game is not a type of crazy golf > it's just a matter of picking the right search algorithm.
