Background
Recently I have been working on a couple of WAM Building Projects. The first being a portable Gunge Tank which I've posted details of in a few other places. (Most recent post about test containing messy crossdressing images: https://umd.net/forums/elise-faces-messy-revenge-after-gunge-ta ,if however you don't want to see messy CD content then here is a link to my video of my first tests using water and nobody in the tank: https://umd.net/videos/prototype-mobile-gunge-tank-initial-test).


Pie Launcher/Device
Anyway besides that project I was also trying to work on a parallel Pie Launcher/Device project. I attempted a couple of different prototypes for this, however each either had problems or fell short of the standard I'm looking for. It's now going to be awhile before I can attempt new designs or try improve on the other ideas I tried but I thought that I might throw the problem open to the community to see if anyone else had better ideas/had experience building something similar/had more engineering knowledge so that this knockoff Dr Robotnik might be able to shortlist some ideas to try at the next opportunity.

I'll first detail essential and desirable requirements for the project then briefly outline what I've attempted already.


Essential Requirements
1. Pies must rest on the level until they are launched, don't want the filling/cream to pour/run off of the pie. I expect this rules out like a large barrelled spud gun/pie mortar for if the pie needs to sit at bottom of barrel and the barrel tilted to aim it other than straight up that the filling/cream will pour out.
2. Must be able to deliver a pie (roughly 500g in weight) into the face of a sitting or standing target. This rules out something like a pie guillotine that requires lying under it and dropping pie straight down.
3. Pie must be orientated to hit target with the creamy side first. Can't have it rotating round and hitting with pie crust/tin first. One suspects this means something like a trebuchet will only work at something like point blank range as the top edge of the pie will be moving faster than the bottom edge such that any pie launched from it will have some over spin imparted on it.
4. If launching mechanism would need to enter into the frame of a close up video if the scene was recorded then it must be able to retract quickly to not obscure the reaction shots.
5. Should not be so overpowered the force of pie hitting may cause injury.


Desirable Requirements
1. Could be used with the new Delivergoo mobile gunge tank I built. People need to be sat rather far back into the tank so trebuchet design requiring point blank range loses further desirability points here.
2. Be as easy to store and transport as the Delivergoo Gunge Tank.
3. Able to launch pies from a distance. No more than 2 meters away.
4. Pies can be automatically be reloaded. (Had in mind either like a makeshift conveyor belt dropping pies onto the launch pad or something similar but this is besides the problem of launching a pie and can be tacked onto many designs).


Designs I've attempted from most simple to most complex
1. Pie Slide: Having a short piece of plastic or wood suspended from roof of the tank that extends slightly out the front in which pies sat on top of the tank can be pushed onto the slide to slide down towards the target. This turned out better than expect (thought it would be complete crap) but pies are not quite getting enough forward momentum to get a full splat or drop a little too short and whilst the pies do have a small amount of over spin as they fall the pie angle when they splat is just off of being a good angle.

2. Pie Swinging Arm: An arm made out of PVC pipes and rotating joints that starts with a pie resting on top of the tank but tries to swing down to sort of hurl it the short remaining distance towards target. Not very good in the end, the sort of wrist joint that was meant to allow the pie's orientation to rotate during the falling swing often spun the opposite way and even if that was fixed this design seriously fails Essential Requirement Number 4.

3. Pneumatic Cylinder powered Pie Shot Put: Pie launch pad on the end of a pneumatic cylinder powered by an air compressor. Various issues came up, far too many leaky connectors especially around the valve. Various cylinders tried but none tested came close to firing the pie anywhere close to the desired height of even a seated person.

And that about covers all requirements and designs attempted so far. If people have any ideas I would be keen to hear them.

Sounds like what you need is one of those sheet-cake-flingers one of the European talent shows uses to eliminate contestants. The cake sits flat on a paddle, which is then catapulted up and forwards (presumably by compressed air), I think the paddles rise to about 45 degrees but the cake is vertical when it reaches the target victim. The paddles are a fair distance back from the target and so don't block the view of the arrival, they also film from both straight on and the side for maximum effect.

Thanks for the suggestion. If you mean Beat It then unfortunately not as the attached screenshot shows. The pneumatic catapults (can just about make out two pneumatic cylinders attached to the bottom) there have the same issues as the usual trebuchet design.

By the time the cake reaches the target (based on screenshot and knowing shows like this use wide angle lens that make studios appear larger than they are in real life let's say from release point to target is about 1.5 meters) the over-spin means it is horizontal and upside-down. Whilst this is fine for the cake here as it has no crust or pie tin and has less to slop out whilst upside down if repeated with a pie it would likely be tossing half empty pie crusts/tins crust/tin-first into the face.

Furthermore the camera angle shows that the pie catapult extends to a full 90 degree angle and well above the height of the contestant fully obscuring any front-on camera angle. Watching back the videos there is no direct front on shots of the cakings.

So, the best pie delivery mechanism by far is the human hand and arm.

If you analyze the motion, it is a follows. The pie is held flat in the palm of the hand. There is enough of a grip to hold it securely. Shoulder, elbow, and wrist joints can adjust aim and propel the pie forward. The torso can provide additional torque to propel the pie towards its target. Feedback loops allow the angle of the pie to be adjusted as accelerates towards launch and allow the pie's acceleration to be carefully measured.

Your mechanical mechanism will need to hold the pie flat, begin to accelerate it, position the pie at the perfect position relative to its target, gently tip it. And then the pie must be accelerated forward with enough force to allow it to reach its target. Too much force will cause the pie to disintegrate and may injure your target.

Unless you're into advanced robotics and heuristics, the mechanism should be simple.

I would think that your first design has the most promise. What about a transparent plastic slide? The pie would slide down until it hit a perpendicular barrier. This would cause the slide to alter its angle slightly (either a sensor or a shift caused by the weight of the pie). This would point the pie at the target's face. The shift in angle would trigger a spring-loaded catapult which would then hurl the pie towards the target's face. The catapult should be hinged such that the flat surface of the pie hits the target evenly.

As a bonus, triggering the catapult would also release the wires holding the upper end of the slide up and it would descend below the chest of the victim allowing a completely unobstructed view of the pie hitting. This release mechanism would moderate the force of the pie.

Can't wait to see what you come up with!

I'd argue that the hand and arm is only the best mechanism for throwing a pie right now as there hasn't been the degree of necessity to outweigh the costs to develop something beyond the likes of a pie trebuchet at near point plank range that is good enough to satisfy most TV producers that want something of that nature.

For instance I've bowled leg spin and off spin in cricket for well over a decade now. 20 years ago if a batsman wanted to practice against a spin bowler in a net they would need to have brought a spin bowler to bowl at them as there was no substitute that could replicate it. Nowadays they can roll out a Merlyn bowling machine that can replicate any spin bowler that ever lived all day long and if desired can bowl with more spin than a human can generate and remain more accurate over a longer period.

Now naturally I don't have the R&D budget the people who came up with Merlyn does but the problem is less complex. If it requires advanced robotics then fine as a software engineer I'm not daunted to try it. However if a more simple solution can be found it would be preferable as with less things that could go wrong the more repeatable the motion that can be created.


Original Intent and Pneumatic Pie Shot Put

My original intent was to create a pie machine that could reliably create the repeatable action of throwing a pie over a given distance (hence why I started with a pneumatic solution) and whilst I was able to create a pie launch pad that would take a pie resting on the flat and tilt it to the right release angle to be launched with no spin that could be mounted on pretty much anything that can produce a linear force the only problem was the pneumatic cylinders weren't producing enough linear force to propel it far enough.


Pie Slide Idea Query

Though I've since relaxed the requirement for launching the pie over a distance if I'm able to incorporate it into the likes of the tank. Hence the pie swinging arm (which turned out to be crap) and the pie slide (which looked to be surprisingly better than expected).

Anyway are you able to produce some sketch of the design for the pie slide you have in mind as I'm having a little trouble following the text description, in particular why it would be necessary for the slide to drop down. If you will excuse what is possibly the worst 3D modelling I've ever done in the attached picture but am in a bit of a hurry but the proportions are about right (though the white cylinder representing the pie may not necessarily be following the exact dropping path) and roughly illustrates the original pie slide idea I tested.


My initial pie slide gunge tank modification idea and tests

In the initial idea I had for the pie slide (red plank here) was fixed and ended above and in front of the tank's occupant such that the pie's momentum down the slide should carry it the rest of the distance in the x plane to the target and the pie's own spin from toppling off the end of the slide winds up in a roughly acceptable angle by the time it reaches the target. With the fixed pie slide the pies can be stored on top of the tank and then pushed off the end onto the slide possibly allowing for like a mini conveyor belt that could hold up to three pies to push on to the slide.

The two problems with this solution revealed through testing was that it has a small margin of error that may mean it doesn't work well for people whose height varies too much from the height it is optimised for or if they choose to lean forward or backwards a bit too much and secondly whilst it might be able to get to the target the momentum in the x-plane means we're looking at more like a pie being lightly placed in the face rather than a good splatting.


Alternative pie slide gunge tank modification idea

One possible solution to this second issue could be instead of a fixed pie slide the pie slide is instead a seesaw and that the slide is extended further away from the tank's occupant with the pie being stored on the very end of the seesaw. Chains attached to the far end of the seesaw could extend up to the winch system that operates the gunge tank to ensure the seesaw can't rotate further away from the occupant than when horizontally level but when the winch is rotated in the opposite direction to operating the gunge tank it raises the end of the seesaw with the pie on and then the pie has a longer slide to slide down and pick up momentum for more splat. It doesn't however solve the problem of people of different heights or leaning too far forward or back and also means that with the pie needing to start at the end of the seesaw that the multiple pies on a conveyor belt option is out though I think most would agree 1 good pie hit beats 3 poor pie hits.

I mean kinda sounds like you're looking for a Pie Pod armature.

Unfortunately not. Both the WWYD Pie Pod and the Beat It cake catapults I analysed as part of my initial research that lead me to eliminate a trebuchet design. Even at its near point blank range the pies often spin out of orientation and the catapults themselves from the WWYD clips I was able to analyse were horribly unreliable/inaccurate with many pies either just falling weakly to the floor or missing out right. (In this regard Beat It's pneumatic power rather than WWYD's elastic power makes it more reliable/accurate). At best the pie pod is about 40% accurate for which the producers try to get around that problem by have 4 or 5 of them so odds are at least one and maybe two get a reasonable hit in. However I've better pie throwing accuracy than 40% such that recreating a pie pod catapult rather than just throwing them from off camera is just unnecessarily wasting pies. Granted making a perfectly accurate pie thrower is a tall order but if I can get something around 80% accurate then it would be an improvement on the probably 60% that I can throw from off camera.

Have you considered an accordian armature that smashes the pie rather than throws it? To keep the pie from sliding off, it would have to tilt toward the target as it extends, but that should be doable. It could be spring loaded to shoot out quickly, and motor recoiled so that you aren't left with a hand in your face.

Also it could end up looking very cartoonish, which I would consider a plus.

Do you mean like the cartoon extendable boxing glove like mechanism?

Hmm there might possibly be something to it if I can just work out if I can make it with something strong enough to support the full weight of the pie through extension and how to mount it onto the tank. I shall put it on my list of ideas to try with Pie Slide/Seesaw along with any other ideas people can think of...

I figure that you could mount it on the front of the tank like so. It would be hinged to fall toward the victim as it extends. Both extension and falling could be powered by springs or bungee cords, which ought to be enough force for a full sized pie.

In the event I pursued an extending scissor arm design for a pie mod on the gunge tank I would be looking to try mount one each between the dual upright poles in the front corners of the tank rather than have one going across the entire front of the tank.

Having one going across the entire front of the tank violates requirement 4 of the project that if the device needs to enter the video frame to perform the pieing it must immediately exit the frame. In this configuration it is permanently stuck in the front on frame for both pieing AND obscuring the front on video for gunging. Given the choice between adding a pie device to the gunge tank that obscures the video for the gunging and not adding a pie device at all then naturally I would error towards no pie device as undermines the very point of the gunge tank itself. Better to do just one of those tasks well than both poorly.

Fair enough. The scissor arm would probably work better if it fell to a 45 degree angle of approach instead of 90 degrees. This would drop it lower and out of frame before extension. With a second device, both could be placed at 45 degrees to either side of the victim, out of direct sight, but covering the whole face with pie between them. As for tirgger and retraction, you would simply need a cord attached to the second center join of the scissor arm that holds it in place before extension, releases that held tension with a button click, and then reels that cord back in to pull the device back out of frame. A fishing reel could accomplish each mechanical demand there, and it may be possible to motorize the reel in.

Just a brief update for those that have been following this thread with any desire to see if any of these pie gunge tank modifications work.

Briefly ran some preliminary tests on the pie slide modification idea posted earlier in the past couple days. Based on those tests my suspicion is that an idea like this could work but that I needed to have planned for this idea earlier in the build process. The impression I get from the tests is that the tank needed to be a little taller to allow more time for the pie to spin/twist into the right angle as it falls off the slide. I've attempted to extend the tank's height as much as I think is safe to do so without also having to widen the frame for further stability which essentially translates as a complete re-build.

Was planning to do final tests on this modification and film it to see if it works when complete or if not how close it got for any future builds/demonstrate the concept however did myself an arm injury - actually from fixing the car at the road side whilst on the way to get more parts rather than from any building work itself - that I reckon may prevent me from being able to do those tests tomorrow. If I do manage to do them though I shall post whatever results I get. If there hasn't been any updates by the end of Monday consider the tests not done.

Also just as a heads up as my previous posts regarding gunge tank builds resulted in a bunch of DMs if you do fire off a message please be patient as it may take me awhile to respond - both due to the arm injury and as I'm also taking some time off the UMD to consider my level of future involvement within the WAM community or continue my kinky nomadry and go explore other things.