Development of a Dynamically Tuned MURLIN Trebuchet

Development of a MURLIN trebuchet which is tuned to work with a bending structure.
Plans are available, (though the non-dynamic tuned version)
at tucsonpumpkinto...
This is moderately challenging to build, and should not be your first woodworking project without help.
The Syrian rebels appear to have built their own version. See • Syrian Catapault

[@shawniscoolerthanyou]

I will never understand anything the way this dude understands trebuchets.

[@PopeGoliath]

This was a thrillingly fast-paced tour through the mind of a maniacal genius. I wish more videos would get to the blazing point like this.

[@FreddyFlinstoneify]

I forgot to compliment you. The journey of your engineering search was very entertaining. Thanks for sharing all of that.

[@ldvance01]

trebuchet dimensions scale well. You also want to keep the CW/projectile ratio about the same and things should work fine.

[@petertimowreef9085]

You often went way too fast for me but I don't mind, it's a video, I can just rewind and watch it again untill I get it. So much better than when the information density in a video is too low. You can't speed up or skip parts or you'll miss things, so it becomes so boring and hard to concentrate on. Thank for making this video Sir.

[@breannestahlman5953]

Very good video. A big thank you back in space / time!

[@ldvance01]

Ketmaniac: Dynamic tuning means incorporating structural dynamics into the design optimization process. In more plain (queen's?) english, this means that the trebuchet is designed with frame bending and cord stretch in mind. The trebuchet is designed so that the energy in the 'springiness' of the frame rebounds back into the projectile before release. It adds about 10% to the efficiency of the throw. LDVance

[@jimmypge]

This is a fantastic exercise in how to approach problem solving, I really love the simplicity of your solutions for getting better real life data on the materials!

[@ldvance01]

In general, as the machines get larger, higher performance materials are required. This is in general true for all structures

[@anachronisticon]

Fascinating. Does this imply that a murlin trebuchet wouldn't have been a viable improvement in medieval times? Improved efficiency/vs fragility or complexity tradeoffs?

[@k1ng5urfer]

@@anachronisticon Since you asked a 6 year old comment a question, here's a 3 year old answer haha. The Murlin would not scale well during medieval times no. Here's why: Out of available materials they would be limited to wood as the strongest available arm material. It was not until the 1600s when early steel starting making a real appearance in quantities that might be sufficient to have used. Even then though, casting of steel wasn't really a thing for another 200 years, nor was forge welding such a huge object even realistically possible. And regardless Trebuchets were phased out some 150-200 years before blister steel was even readily available. With only wood to use, even the strongest materials like oak, ash, or fir would not hold up to the kind of loads that they would sustain on a Murlin. At least not when scaled up enough to throw any weight big enough to make it an effective siege weapon. Additionally lack of strong/flexible enough ropes and cables would likely be another deal breaker for assembling this thing in medieval times.

[@TsadikTsadik]

Love your work mate, I am getting ready to build a murlin this school holidays with the kids.

[@ldvance01]

Tsadik Tsadik Glad to help with plans - send a request into the supplied email address and let me know you're OK with the normal two conditions

[@ayeletfine5420]

Eco contact me lost your number what's app me ayelet we r in Queensland

[@ickotsa]

Ayelet Fine Ayelet, I will call you soon. my mobile died, so when I get my new mobile I will give you a call.

[@ayeletfine5420]

Ok darling

[@binary841010100]

damn !!!! he did all the calculation... amazing dedication

[@RobotsEverywhereVideos]

Dear sir, you are the king of the geeks. Please go back in time and take over the world with your awesome siege weapons.

[@yardsaleuw3075]

Your throwing arm measurements should match the Febunaci sequence. Constant acceleration will only happen when the driving rope material has zero stretch.

[@ldvance01]

Yes. I optimize all the sprocket lengths along with the throwing arm and sling using a genetic algorithm with a simulation. Slight variations on a baseline are chosen, a throw is made, and if it throws further, then it is saved as the new baseline. Repeat that a couple of hundred times, and you have an optimized trebuchet.

[@npscull83]

This is very interesting. Our ASME club recently built 2 trebuchets, our team built an f2k and we threw pumpkins and bowling balls about 100 yards using over 700 lbs of counterweight. I wish we would have had time for dynamic tuning like you did here, I bet we could have made it much more efficient. Floating Arm Trebuchet (F2K): EWU ASME Club There was a MURLIN design at our competition but it didn't do very well. Do you have any experience building MURLINs large enough to throw pumpkins in the 100+ yard range?

[@chrisbledsoe1970]

Sir, we’re planning to build a machine such as this for a pumpkin chunking competition if you still see these comments I can’t seem to find plans anywhere for one of these. I want to modify it a bit and instead of using “sprockets” use a cam similar to the ones on a compound bow. I want to do this to help eliminate the bounce seen in most of the Murlin trebuchets. Would you be willing to help us out? Thanks In advance for your time.

[@qumpania]

genius ! amazing presentation, well done !

[@FreddyFlinstoneify]

At 125mph launch speed at 45 degrees angle, with gravity being the only external force, the distance should be about 1050ft. At that same speed at 30 degrees, the distance is somewhere around 900 ft. I understand that you are looking for energy loss. Is distance important in your design? If so, should the launch angle be closer to 45 degrees? My simple calculations assume that speed difference is negligible between the release at 45 and the release at 30. Thoughts?

[@warmfreeze]

plans please? i want to scale this down by 70% to throw marbles. I lose my marbles often however i keep finding them again... i want to lose them once and for all!

[@ldvance01]

+warmfreeze : I assume you found me on the proscribed email location. Instructions for getting plans are in the annotations for the videos.

[@ThatGuy-ev2cb]

+Leonard Vance I don't see the link for this project. I would like to use your model for a school project, so if you are willing to, can you send me plans? As for updates on the project, I am required to take photos along as it progresses, so I can send you the photos as my group builds.

[@equityplus]

find a way to add backspin to the ball when released. Although you might argue that energy is used in adding spin, the flight characteristics of the ball will greatly improve and the launch distance will improve. There are tons of information on spin and ball flight for golf balls. I know you are in the hunt for lost energy but this will give you more distance.

[@ldvance01]

+equityplus You are right, and I separate the two ends of the sling so that topspin is imparted on release in order to maximize distance. The THMPR trajectory actually launches low and curves up significantly, just like a good golfball drive. High performance pumpkin chunkin trebuchets actually tone down the topspin as the pumpkin can curve up so high that it actually loses distance.

[@FingerAngle]

You meant "backspin", not "topspin". Topspin drives the projectile into the ground. Yes tuning the amount of backspin is important.

[@FingerAngle]

Mr Vance just miss spoke.

[@FingerAngle]

I like using magnus for line of sight targeting. I have video of beautiful magnus shots from one of my trebuchets. I've been studying magnus effect for several years, and how to control it.

[@tommylum]

Great commentary on how much went into this "simple" machine. Can you run down the stats for me? 93lbs of CW? Baseball for a projectile? 164 cm drop of CW? What length is your throwing arm from tip to center pivot? 813' of distance? Just want to compare to similar trebs...

[@ldvance01]

My current opinion is that MURLINs and F2Ks (a FAT with an initial vertical drop before engaging the horizontal track) are roughly comparable high efficiency trebuchets. The standard MURLINs hold good performance up to mass ratios of about 400, but I don't know how high you can push an F2K. Maybe an other builder can say.

[@nomercadies]

Offered in the most friendly and sincere tone. Enunciate more clearly and slow down the delivery. I believe those steps would increase the understanding of your data from 80% to 99.5%. You had nearly half the video time left as it went to black half way through the thirteen minutes plus you uploaded.

[@ldvance01]

Trevor: your pumpkins will be about 50times heavier than a softball, so you would have to beef things up a lot, and push the CW up to several hundred pounds. If you are still game, let me know if you're OK with the two normal conditions, and I'll send baseball plans over.

[@COeng0612]

Thanks for the help!

[@COeng0612]

Great video! I am interested in getting your basic plans for this trebuchet. I am more than willing to agree to the terms that have been listed previously in this thread. Would you be willing to send me the plans?

[@ldvance01]

I'm afraid trebuchet dynamics really cannot be expressed without such 'university talk'. The underlying mathematics are simultaneous nonlinear differential equations. This particular video goes through a pretty typical cycle of hypothesis, prediction and testing common to most research, and is meant for an engineering audience.

[@DrippDroppz]

@Leonard Vance @EVERYONE , My question is, if you want to increase distance of the projectile and efficiency , why not remove the pulley ? Would the drag force be less gliding along the wooden dowels rather than a pulley? Please help me understand

[@ldvance01]

@ACDC5968: Hmmm... Were you thinking of trying it on a different kind of trebuchet or on a smaller MURLIN like the golfball model? LDVance

[@breannestahlman5953]

Is murlin more efficient than floating trebuchet? Is it the most efficient?

[@ldvance01]

I'm saying something close to: "frame bending and cord stretch". My simulation at the time didn't account for either, and it was making a big difference at larger counterweights.

[@guitarshredder1972]

probably a dumb question but, is there a way you could use pulleys like are used on a compound bow?

[@ldvance01]

I've got to give the other trebs some more thought. The MURLINs are dominated by a single degree of bending, namely the stretch of the CW cord. For configurations like whipper, FAKA and F2K, there might be multiple bending items to be modeled. It might be beyond my math skills. Need to give it some thought. What about First in Fright?

[@ldvance01]

A continuous cam is harder to make, increases the inertia of the throwing arm and doesn't appreciably increase performance.

[@FreddyFlinstoneify]

@ldvance01 - I get the same. Thanks for responding. Out of curiosity, have you evaluated the spin produce from a sling? If you have, I am very curious what you have learned and how you evaluate the proper angle from that. Again, thanks for responding.

[@ldvance01]

@FreddyFlinstoneify: 45 degrees is the optimum distance only in a vacuum. Aerodynamics dominate the flight of a baseball at these speeds, with both drag and lift (from spin - see Magnus effect) being important. Lower trajectories are optimal when drag is present. zero drag distance is actually close to 1880feet if I use 75m/s as an initial velocity - see if you get the same. LDVance

[@385machine]

Amazing video! What were the counterweight mass, projectile mass and machine dimensions were used for the 2008 and 2009 competitions? I made this type of machine and recorded a 22 meter distance throw with 1.38 kg counterweight and 50 g projectile. The max machine height allowed was 65 cm.

[@The_Conspiracy_Analyst]

You should try dacron BOWSTRING, this will drastically reduce energy losses due to stretch (exact reason it's used in bows)

[@stingray1227]

I have a few questions: 1. Will an about 1m tall version launch a tennis ball over, say, 30 m? 2. Does this use any pulleys, elastic forces? (is it solely powered by gravity?) Please respond. Thanks

[@larrychristie5805]

i would like to enlarge the murlin to chuck 8 lb pumpkins! Has this been attempted with any success? I have access to a variety of materials such as aluminum poles, tractor weights etc I agree not to distribute the plans and to pass my results on to you

[@para3668]

6:52 my "Nice" reflex kicked in

[@ldvance01]

A smoothe spiral would work pretty good. It's just harder to build and produces a throwing arm with (probably) higher rotational inertia, which ultimately reduces efficiency

[@steffankaizer]

put 2 tiny nails in the "poles" of the ball and you can ditch the pouch for 2 tiny loops. is it more evitient than a floating arm?

[@gs7schuler]

I have seen your videos that my friend showed to me and we are very interested in building one. We just built a floating arm trebuchet and have made some small models of all different kinds. I was wondering if you could send me the plans so we can take out trebuchet building and throws to the next level. Thanks

[@SUPERNOVASPECIAL]

Have you tried a cam system rather than the five arms? Maybe that will mitigate energy loss. If I'm wrong I'd like to know why. Any thoughts?

[@ldvance01]

Zach: A continuous cam helps because the forces don't cycle up and down as much, and this adds about 5% to the efficiency, because energy is lost in the stretch cycling. There's efficiency loss, however in the increase in throwing arm inertia, so it's best to make the last two sprockets continuous because that's where the forces are the highest, and inertia increment the smallest

[@axelbengtsson8195]

Is vertical weight energy transformation harder to apply on larger structures, since the structure and rope are subject to more strain?

[@error.418]

Traditional trebuchets saw in increase in efficiency and throw distance by mounting the platform on wheels. Have you considered this approach, or does it not fit the mechanics of this particular system?

[@ldvance01]

Wheels are useful on traditional hinged CW trebuchets because they permit the counterweight to drop closer to vertical, permitting more energy transfer to the projectile. The MURLIN drops the counterweight vertically to begin with, so wheels would just reduce efficiency.

[@error.418]

Leonard Vance Thank you!

[@48ford8n]

The hidden loses are coming from "interaction" terms which cannot be found since all your experimentation is OFAT "one factor at a time". If you employ statistical design (DOE) experimentation on the key factors, you will be able to find these last remaining loses.

[@ldvance01]

+D cirval (48ford8n) Well, if you're right, there's only about 5% hidden.

[@IronGoober]

I do not know that much about trebuchets, but has anyone tried a spiral instead of a set of nodes? It should result in a smoother acceleration of the arm than a set of 5 nodes...I think. Any reasons it would work less well than 5 nodes?

[@drewm6925]

Sir, Loved your video! I agree to your terms, would you please send me the plans to build your design? I am building one with my son and his friends for a high school physics project. Our projectile will be a golf ball. Our goal is to launch our projectile at least 100 yards. Thank you for your consideration.

[@Mr400babies]

When are you making one to fire bowling balls? Would it be listed as a destructive device?

[@ldvance01]

Kieran Moore Kieran: Take a look at the American Chucker channel or facebook page. You can see our current attempt to challenge Yankee Siege. This may not be practical for the home builder though.

[@ldvance01]

@FreddyFlinstoneify : I haven't evaluated the spin more than qualitatively. It apparently doesn't take a lot of spin to produce a substantial lift (see Magnus effect) I"ve guessed at the right lift coefficient and optimized for launch angles, getting about 35 degrees. LDVance

[@WeBeTom]

LD, I keep coming back to this video because something was bugging me. I think maybe now I know what it is; the sling. Is the sling used on basballs in this video the same style as that used for the GolfBall Challenge? Is it possible that the release experianced with this kind of sling adds resistance to the ball's velocity? I'll open a new dialogue @ thehurl.org. Walker

[@ldvance01]

Thanks very much. I'll take a look.

[@ldvance01]

Good advice, I'm still experimenting with this movie thing.

[@ldvance01]

Well, it really IS complicated. Even without compensating for effects like structural dynamics and rope stretch, trebuchet mathematics are still multivariable nonlinear differential equations. But sometimes, using all that complex math, an designer can figure out how to make something simple that works really well. It's hard to do, but something all engineers should strive for. LDVance

[@yegermister1]

What program did you use to to model the trebuchets movement?

[@ketmaniac]

@ldvance01 Ah, thanks. Yes, looking at trebs of all kinds here on KZbin, I'm always thinking they're too tall for their base. A lot of constructors don't seem to appreciate the size of the horizontal forces, all the violent pushing and pulling that goes on. Their structures are well able to withstand vertical forces, but horizontal ones not so much. The bending and sagging involved absorbs quite a bit of energy, I suspect. I'd be interested to know what you mean by "dynamic tuning"...

[@ACDC5968]

@Idavance1 you should try doing your dynamic tuning to different trebuchet

[@TheMctp17]

I am trying to scale this down to shoot a 2 lb ball 50 yards but also be able to launch the ball as far as possible. Is it possible to adapt your baseball murlin plan design to shoot 50 yards accurately? Could this be done with adding minimal weight? thanks

[@ldvance01]

the 2 lb ball is about 6 times heavier than a baseball, so you will have to cut down the length of the throwing arm (not the sprockets) so that the arm doesn't stall prematurely. That said, you should have the ability to easily get to 50 yards. Accuracy is much the same as all trebuchets, which is pretty good in general

[@Varue]

thats pretty good efficiency

[@WeBeTom]

Great presentation, LD. This is the kind of work I love. I wish I could be even 10% as scientific but I'm just not designed that way. Walker @ the Hurl

[@ldvance01]

90 lb counterweight(+ fastener HW), baseball projectile (144g), with a 7g pouch and 164cm of CW drop before release. main arm length is 85.6cm and release velocity is 75.1m/s, giving 62% effeciency (I think I screwed up and said 69% on the video). Distance is 248m or 813ft, with a modest tailwind. I don't look too closely at distance these days, there's too much uncertainty in the wind. LV

[@ShrpBld78]

You said that you adjusted the rope length to take advantage of the full distance you had to fall. If I elevated the trebutchet and adjusted the length of the sprockets so I could use a longer rope to fall further than the base of the trebutchet do you think that I could get more performance out of it?

[@ldvance01]

+ShrpBld Probably yes. If you proportionally increase the node distances by the fractional increase in CW drop height, it should work. There's a limit to how far you can go because eventually you increase the inertia of the throwing arm when the nodes get larger, but for modest changes I expect it would work.

[@ldvance01]

+ShrpBld Yes, that would improve performance. Local pumpkin tossers here in Tucson have done exactly that, putting the treb up on blocks of wood so that the CW can fall further. I suspect they did it because the CW became taller, and was hitting the ground though...

[@ShrpBld78]

+Leonard Vance Thanks for replying back, I think that we will work that into our design. I have been trying to think of a way to design our nodes to minimize the amount of weight. We were thinking of making the throwing arm like a wood I beam that gets smaller at one end and rapping in a layer of fiber glass. I was thinking of using pvc for the nodes as it is a light cheap material.

[@ldvance01]

+ShrpBld PVC is not a structural material of any note, may not be strong enough to hold the CW

[@brendandor]

cool video though why is the audio only out oneear and half the video blank?

[@ldvance01]

I don't know about the audio, but the video length was simple incompetence on my part. I'll take a look, thanks.

[@ldvance01]

are you good with the normal two conditions?

[@ketmaniac]

You directed me here from your MRT vid. This is a MUCH better machine. But I have to say I took a long time viewing the vid. I'm not American and I had to keep replaying bits to try to decipher your high-velocity speech. 1:06 - "It took me a while to figure out....huge amounts of energy". The middle part of the sentence is lost, but it sounds like something crucial. What are you saying at that point?

[@charleshealey4909]

Hi, I have been watching a lot of your Murlin videos and I'm interested in building and the physics behind trebuchets. Would it be possible if you could send me your plans to the Murlin trebuchet?

[@10babiscar]

why don't you use a cam instead of sprockets?

[@wolterh6]

Did you optimize the lengths of each of the lever arms?

[@thatoneguy99100]

well done sir

[@kingkill666555]

Anybody know the program used here for the model running?

[@brandysigmon9066]

Would you lose some energy through the friction of the spinning pieces?

[@ldvance01]

The energy loss from a rotating bearing (as measured) appears to be insignificant compared to rope bending and aerodynamic losses. This was somewhat of a surprise when I first did the calculations, but I verified these results with the rope loss experiment I did in the middle of the video.

[@ACDC5968]

@ldvance01 no like different types of trebuchets like whipper, faka, and F2K

[@conheo1992]

what kind string to hang the weight

[@ldvance01]

I use MatLab - but the simulation code is my own and not currently available - sorry.

[@ACDC5968]

You sound just like my pop pop he knows all about that math stuff and he loves making trebuchet.

[@ldvance01]

My apologies - I thought you were connected with the "First in Fright" Crew. Not sure where I got that impression, come to think of it... LDVance

[@charleshealey4909]

I also agree to not distribute the plans and update you on my results

[@karenstalick2868]

can you send me your plans. I would like to build one for my physics project nicks

[@ACDC5968]

@Idvance01 what about first in fright.

[@ldvance01]

I do share the plans, though under the normal two conditions. If you're OK not distributing the plans any further without my permission, and letting me know how the building and testing go, send me your email address and I'll pass the design package on.

[@rjwintl]

Physics !!! yeah, buddy !!!

[@dodgerules22]

I am up to game and i plan on letting you know right away!

[@meganpainter5388]

What software did you use to make the simulation?

[@ldvance01]

I do all my simulation work in Matlab.

[@ldvance01]

Sir: I give out the plans for golfball and baseball 5 armed MURLINS, though not dynamically tuned. Still good for 700-800 feet. Conditions are: 1) you agree not to distribue the plans any further without my permission and 2) You let me know how building and testing go. If you're OK with the conditions, leave me an email address and I'll send the package over. LDVance

[@ldvance01]

@gs7schuler : I give the plans out on two conditions: 1) You agree not to distribute the plans any further without my permission, and 2) you share your results as you build and test the device. If that's OK, leave me an email address and let me know whether you want the baseball or golfball device. LDVance PS: I am giving out non-dynamically tuned plans - performance is still pretty good though...

[@ldvance01]

Performance and efficiencies of the MURLIN are comparable to those of the F2k, at least qualitatively as far as I've seen.

[@SpiridonovRU]

Greetings from Russia and thank You for video and great job

[@FreddyFlinstoneify]

Last post, I promise. I thought the launch angle was around 30 degrees from 5:50 point in the video.

[@polimecanos]

Hi there, did anyone notice that the video ends at 7: 26 and the remaining time to 13:15 its a blackout. besides that, thanks for sharing nice job.

[@ldvance01]

polimecanos Yep - that was my error. Anyways, it's up there and working, so I've never reloaded it.

[@bradharris3323]

Dimensions?

[@ldvance01]

@nanchulantturtle: Sure thing - the plans I send you won't have the dynamic tuning, but you can still get 700 feet from it. There are two conditions: 1) you won't distribute the plans to anyone else without my permission, and 2) you let me know how things go as you build it, since I use your feedback to improve the design package. Let me know if you're OK with these conditions and send me an email address. LDVance

[@mangotango2481]

what is the name of the simulator that you use?

[@jonahhadley22]

matlab

[@billkratzer1]

excellence

[@Bravo872]

Your CW needs more time to accelerate.

[@pulsefrequency]

amazing work. genius kids.

[@rjwintl]

build one for the local high school to be used @ football games to throw effigies of the competition literally "out of the stadium " !!!

[@moflyboyblanquito541]

Coriolis?

[@ldvance01]

Can do. Can you send an email address? You may have to try more than once as sometimes KZbin buries some such comments under the spam folder