Hello all,

During the preparation of the exhibit 'The Sword - Form & Thought', Peter Johnsson and myself worked on a tool that made it easier to document and display the dynamic balance of swords. We are finally ready to share it with the community at large!

You can read more about its context and principles on this introduction page. Then you can use the tool itself, with the help of the more detailed walkthrough.

I hope many people will enjoy working with it! If you have any question about it, do not hesistate to ask.

[ Linked Image ]

Best regards,

I have been eagerly awaiting this! Thank you!

All vendors and manufacturers should provide a cart like this with every sword.

The vibration nodes you are putting on the diagram are in the plain of the flat. Which is less important than the vibration nodes in the plain of the EDGE I believe, when it comes to force transferred in the cut. Though the nodes for the flat are extremely important.
The only person I have heard mention this is, Matt Easton.
But Is there an accurate way to measure the node in the plain of the edge, Other than hitting a piece of wood and feeling where it is?
I would really love to see this in the diagram. :D

I will use this on everything. Expect graphs of crowbars, spades, LARP swords and, if they stay stationary, cats. Just for reference of course.

I have actually thought about edge to edge nodes a little bit :) Some remarks:

• As Timo points out, they are quite hard to measure. Most swords are stiff enough edge to edge that the motion will be tiny, and any sound made will probably be masked by the flat-to-flat higher modes. It is very hard to stimulate only edge-to-edge vibrations.
  
• Functionally, I'm not sure they have a noticeable effect. In handling you will never notice the vibrations, and in cutting the vibrations are not detrimental, you only get a small oscillation in the force applied on the edge. Flat-to-flat vibrations, on the other hand, can increase friction or disturb edge alignment (or at least they can be related to these).
  
• I have actually been simulating swords with a finite element method, which bypasses the need to measure. It can be shown with this that the nodes are indeed different in the two planes. However they are often not that different, unless you have a sword with a strange profile (rat-tail tang) or very evolutive fullering. The nodes of the primary mode of vibration are heavily influenced by mass distribution, which is the same in both planes.
  

For all these reasons, I don't think it is useful to focus on edge-to-edge vibrations in the current state of our knowledge and methods.

Best,

Wow, this is so cool! Thanks for that. I guess I'll try it extremely soon :D

I just thought I'd mention it. :) I personally am more focused on pivot points.

Something I thought of is to hold the sword with some sort of callipers for accuracy in the waggle test.

Any tool for doing the waggle test must be capable of taking both fully mounted swords and the naked tang of well preserved originals. And do this without leaving any marks. Such a device must also be so simple in construction that it can be built by all who are interested in taking this kind of measurements in a standardised way.
If everyone will use his or her own tool we still have a situation with data that is not necessarily complementary.

I think that when a tool to measure swords is developed, it would be better to just abandon the waggle test and go for a pendulum test. I have cobbled together such a setup before, I just need to streamline the design in such a way that other can build their own. It would be safe enough to use on original swords.

The waggle test would then remain the gearless, accessible method it already is. People who are unsure of their accuracy, or need more precision, would use the pendulum test. The online tool could easily be adapted to treat measurements via a pendulum test too.

It is all in the pipeline :)

Regards,

I agree, the waggle test should stay the waggle test. It is accurate enough for pretty much everything, save reproduction. ;)

I'd agree with Vincent's view that a pendulum test is the way to go for a reproducible, dare I say 'scientific' approach.

I have a relatively simple set-up that I've been using (see attached). The components (excluding tripod) only cost me about £15 and whilst I got the body machined by a colleague, the only important feature of the body is a straight bore of the right size to take the roller bearings.

This set-up seems to work reliably and the clamp copes well with a variety of sizes and can grip a bare tang without any difficulty. I'm still not sure whether it would be considered sympathetic enough to be used on an original museum piece though - others (Peter?) would have a much more informed view on this?

---

 Attachment: 77.37 KB

Given the surge of interest in the matter, I'm going to share my own pendulum setup sometimes next week. I just have to build it anew :)

Here are the constraints that I've set for myself:

• Low cost, low practical skill needed (if I can build one, anyone can ;) )
  
• Low weight, portability
  
• No calibration needed
  
• Can adapt to any sword (or non-sword) shape, without putting too much stress on the object
  
• Can measure oscillation period from several precise points (just as can be done for the waggle test)

The main problem I have with your device, Simon, as far as I understand it correctly, is that you will need to calibrate your measurements to take the inertia of the rotating part into account. Also, in what way do you measure with precision where the axis of rotation falls on the weapon?

I might be wrong, and we can certainly live with several alternatives for that setup anyway!

Regards,

Those are valid questions Vincent.

The inertia of the gripper is clearly small in proportion to that of the sword but I admit I haven't calculated it yet. As regards the axis of rotation, the design/dimensions of the gripper means that it can be made to fall on the central axis running down the length of the sword in most cases, which is what I wanted, but would work less well with different types of object. I've not yet quantified the measurement uncertainty associated with this issue.

You could add to the list the friction in the bearings that I'm using in the housing, and any error induced by the clamp/housing not being absolutely level - I use a spirit bubble to level the device but there's more uncertainty there.

I will have to put some values to these but I'd be surprised if they are significant compared to the uncertainty associated with measuring the period of oscillation and the simplified physics model being used to calculate moment of inertia, percussion length etc.

I'm currently exploring the potential use of reverse engineering/CAE methods to generate much more sophisticated results, but that as they say is another story and clearly not something that would be practical for widespread/general use.

Waiting with interest to see your set-up,

Cheers