Articulation

[Mac]

I am starting this thread to serve as a way to record and present what I think I know about the articulation of armor. It is my hope and goal to cover all aspects and types of articulation. My intention is to start out quite simply and to add more material every few days until I have amassed a sort of treatise. This will probably take a number of months, so I ask for your patience and kind indulgence.

Please feel free to ask questions. I'm sure that I will say things that need clarification, and welcome any input that will make this thread more valuable. Let me apologize in advance for the number of times where my reply will be "we will get to that later".

Let's begin with the very simplest sort of "pivoted" or "hard" articulation. This will involve two plates and two rivets. The goal is to achieve some degree of angular deflection between the plates while maintaining a sort of structural and defensive integrity.

I'm going to begin by modeling this in cardboard. I expect to move on to two dimensional ways to represent this in the next post.

Here are two strips of cardboard about 7 1/2" by 2" (180mm by 50mm). They might represent lames, or they might represent larger plates. It is only the articulating interface that concerns us right now. The locations of the pivots have been punched at about 5/16" (8mm) from the edge of the overlapping piece. For our purposes, the other piece will underlap rather significantly. The reason for that will become clear later.



Here we can see the two pieces have been curved into approximate semicircles and joined with paper fasteners.



The "horizon" of the assembly is more or less straight when compared to a ruler.


If we attempt to flex of collapse the assembly, there is only the slightest angular change when view against the ruler. This should surprise no one.



Now, let's cut some of the overlapping edge away.


When reassembled, we find that the assembly is still straight when in its extended position.



However, we can now make the assembly collapse or flex to some degree.



Since this was clearly a step in the right direction, we will try cutting a bit more away.



The result is more motion.



Let's do that one more time, and cut a bit more away.



The result is even more motion.



Depending on the project, this may well be all the motion we need, and for now let's assume that it is.

Let's turn our attention to the underlaping plate. The curved lines represent where the edge of the overlapping plate stopped through the various iterations of trimming and flexing. The space between the edge of this plate and the last of the curved lines is our final underlap.



I've drawn a line parallel to the last of the curved lines to represent our desired underlap. Note how the line swings away as it nears the pivots in order to leave sufficient material for strength. The shaded area is what we don't need or want inside our assembly.



... so it gets trimmed away. Variations on this shape for the underlap are seen on authentic armor. In general, the higher the quality of the armor, the greater attention this receives, and the closer it approaches the ideal.



This simple structure is the foundation upon which all "two pivot" or "hard" articulation rests. I know that sounds like a bold claim, but it's true. I will elaborate on this in future posts.

Next, I plan to show how to model this joint in two dimensions. While that may seem like a step backward, it allows for very convenient planning, and helps to understand and specify the parameters of the articulation.

Mac

[Carter]

Really cool to be reading and have that sudden moment of realization at how it connects to an armour piece.

How do you determine the curve of the template? I'm sure your not using truncated cylinder's given that few articulations are perfect cylinders if they're even circular at all.


Is it just familiarity and guessing?

[Mac]

Really cool to be reading and have that sudden moment of realization at how it connects to an armour piece.

It's all about seeing things in different ways and using that information to understand what needs to be done.

How do you determine the curve of the template? I'm sure your not using truncated cylinder's given that few articulations are perfect cylinders if they're even circular at all.


Is it just familiarity and guessing?

I typically extablish curves by guess and by eye. That's close enough, really.

The only time anything in armor has to approach being truly circular is in places where the parts must rotate. Even there, the expectation is to have some slop to allow for deviation from truth. I hope to address turning joints sometime later. If I forget, I hope you will remind me.

Mac

[Mac]

I'd like to talk about modeling articulations in two dimensions. This has a lot of application "on the drawing board", where pencil lead is cheap and eraser crumbs pile up thickly. If the basic form and proportions of the articulation can be worked out in a full sized sketch, it saves a lot of trial and error with steel and hammer. Fortunately, rendering an articulation as a saggital section through its mid line will provide an accurate approximation of what it will do in three dimensions.

Let's start with that cardboard model of the most basic "two pivot"/"hard" articulation from the first post. I've sketched it out in two views here. The cross section is just to extablish that the thing we are working with is fundamentally a half of a cylinder. It's not strictly cylindrical, and it doesn't have to be. I will address different sectional shapes somewhat later. The second view is really the important one. It's a "side view" or "elevation". The underlapping part of the inner lame is shown as a dotted line. The paper fasteners are sketched in, but they probably should have been omitted in favor of a simple dot or cross to mark the pivot.



If we set one leg of a compass on the pivot point....



... we can use the other leg to show the arc which the edge of the outside lame will describe as the parts are flexed. That is to say where the mid-saggital point of that edge will go. We don't need to worry about the rest of the edge. If it works in the middle, it will work everywhere else. That's not quite true, but it's practically true. Let's accept it as true for now, and I will talk about the limitations later. The compass arc terminates where it reaches the surface of the inside lame.



This picture shows that our compass arc makes a pretty good approximation of where the edge will actually stop. The line on the cardboard model was made by running pencil along the edge when the articulation was "stopped" or "locked".



Likewise, the compass can be used to trace out the path of the underlapping edge. (My compass point slipped while I was posing this picture... just pretend it's still on the pivot) The important thing to note here is that the edge moves away from the inside of the outer lame. It moves into the internal space and does not bind. This edge is free and does not impede the movement. This will become important in articulations that we will discuss later.



Another way to make use of this side view involves a light table or light box. I mostly use a traditional one with frosted glass and fluorescent bulbs, but the new LED ones are inexpensive, flat, convenient, and have very even illumination. They are better in all ways, really, but I'm stogy.

Here I show how one of the lames is traced onto a piece of printer paper.



I've traced each piece on a separate sheet of paper because I thought it might be easier to see. In practice, you can just trace the lame you are interested in, and work on the original drawing.



If we superimpose the tracings and pin the pivot point in place with a pencil tip......



....we can rotate one of the papers to simulate the movement of the lame.



This gives us the same information as the compass, but without the distracting arcs. This method will also give us the opportunity to try out alternate pivot locations. This will become very important and useful in some applications, and will be addressed later.



This pic shows that our two dimensional, center-line model matches well with our three dimensional cardboard model in predicting how much the plates will move before "binding" or "locking".



Alternatively, we could cut out these two dimensional, center-line models from paper of cardboard. Note that the pivot locations have been pricked here. If they were cardboard I would have used a small leather hole punch.



By physically superimposing one on the other, and pinning them at the pivot point (here with a pencil) we can manipulate these models without need of a light box. If you were experimenting with the pivot location, you would only make a hole in the upper layer, and use the pencil point to mark the test location on the lower one.



In the next installment, I plan to discuss some of the direct applications of this basic articulation. I'm sure that most of you can imagine some already.

Mac

[Indianer]

Mac, before You continue - it might be possible You are blinded by focus. The direction You are modelling the articulation in only occurs in gusseted elbows (compression articulation). Everywhere else the bow is on the other side. I mean, technicalities still apply. Just thought...before You get too far with this, I'd mention it. Cheers

EDIT: I'm not dissatisfied now Mac^^ Didn't mean to imply that. No I just thought You might be cought in thought. Happens to me all the time...often when I'm welding. Which is bad. Anyway, curious to see where this goes!

[Mac]

Mac, before You continue - it might be possible You are blinded by focus. The direction You are modelling the articulation in only occurs in gusseted elbows (compression articulation). Everywhere else the bow is on the other side. I mean, technicalities still apply. Just thought...before You get too far with this, I'd mention it. Cheers

This is just the starting point. Bear with me, and I think you will be satisfied with the result.

Mac

[wcallen]

Mac, before You continue - it might be possible You are blinded by focus. The direction You are modelling the articulation in only occurs in gusseted elbows (compression articulation). Everywhere else the bow is on the other side. I mean, technicalities still apply. Just thought...before You get too far with this, I'd mention it. Cheers

EDIT: I'm not dissatisfied now Mac^^ Didn't mean to imply that. No I just thought You might be cought in thought. Happens to me all the time...often when I'm welding. Which is bad. Anyway, curious to see where this goes!

Yea, my first reaction was similar, but then I thought about what Mac is posting for a little bit. He is right, it will all become clear pretty soon.
"Stay tuned."

Wade

PS - Thanks for doing this Mac. Someone needed to actually write this stuff down, and you have a much more methodical way about you than I.

[Mac]

My intention for this installment was to talk about applications of that fundamental articulation, but I realized that I had a couple of points that needed to be addressed before I could do that.

The first is something I touched on in the initial post when I kept deepening the curve on the cardboard until I got a decent amount of motion. Now that I have shown that the two dimensional model taken down the mid-saggital line is a good representation of modeling in three dimensions, let me show how that looks on paper.

In this diagram, we have representations of three basic articulations. The "depth" or "height" from the pivots to the "horizon" of the lames remains the same. In this case that's about 1 3/4" or 45mm. The shape of the inner lame is always such that it follows the edge of the outer lame and underlaps it by the same amount. The variable is the distance from the pivot to a line drawn perpendicularly from the edge of the horizon of the outside lame. This is the thing we changed in the cardboard model by cutting material away from that edge.

In the first example, the distance from the pivot to the perpendicular is about 5/16" or 7mm. When you articulate the inside lame on the light table, it stops at about 160° when measured along the horizon. We might also think of that as 20°of useful flexure from the starting position.

The second example shows what happens when that distance is increased to 7/16" or 11mm while keeping everything else the same. Now the angle along the horizon is 150° or a 30° change.

In the third example the distance is 5/8" or 16mm, and the angles are 140° and 40°



As we can see, when everything else is kept the same, the amount of motion in the articulation is a function of the distance of the pivots from that perpendicular dropped from the edge of the outside lame. The greater that distance, the more the motion. Looked at another way, the more curve the lames have when cut as flat sheet, the more motion the articulation will have when they are shaped up and assembled.

In the next set of examples I will change the height of the lames at the mid-saggital plane while keeping the distance from the pivots to the perpendicular to the edge the same. In this case that distance is about 3/8" or 19mm.

The heights of the four examples are shown in the diagram. Like the previous diagram, the articulations are shown in both states and the resulting angles are shown. Also like the previous diagram, the inside lame is shaped so as to have an underlap of uniform width.



We can readily see that the smaller the height, the greater the motion. This is an important thing to remember. Within limits, a shallower articulation will yield more motion than a deeper one. I will have more to say about this later.

The third thing I'd like to address today is about the shape of the articular edge of the overlaping lame. In all the examples I have shown so far, that shape was the one we ended up with by trimming away material from the cardboard model while still keeping enough material to provide the pivots with strength. What I want to point out here is that as long as the two parameters I've just talked about are the same, the shape of the articular edge of the outside lame is on no functional importance.

The first figure in this diagram shows the articular edge shape we have been using, with the two important parameters labeled A and B. "A"is the distance from the pivot to the perpendicular. "B"is the height from the pivot to the horizon, or the point where the mid-saggital plane meets the articular edge. The subsequent figures show five possible shapes for the articular edge of the outer lame. They will all work identically. They are stylistic choices and not functional ones.



Having covered these ideas, I hope to move on to applications of this basic articulation next time.

Mac

[Mac]

Before we can discuss applications of this basic articulation, that is one more point I'd like to cover. This concerns identifying what parts of the articulation are actually "doing the work". That is to say what regions or edges actually define the articulation and produce the "stops" at the ends of the travel.

If we begin with our model in the flexed position, and extend the parts, the underlapping edge of the inside plate swings along an arc until it binds on the underside of the outside plate. This location is marked with a red exclamation point where is touches the horizon and red shading where contact will happen along the surface if the model were three dimensional. This represents that critical area which produces the "primary lock"



In the two dimensional model, we are only able to find the point on the horizon where the lock occurs, but in the three dimensional model, we see the interference along the entire surface. In effect, the lock occurs when we try make the longer curve of the inside piece pass through the slightly shorter curve of the outside piece. This is an important idea to understand. It is not only the foundation of this basic articulation, it is at the heart of all two-pivot or hard articulation which are derived from it.

Here, the important underlapping edge of the inside plate is marked in red.



As the articulation is extended, that edge strikes the inside of the outer plate. That contact happens all the way from pivot to pivot, and further extension is impossible without damaging the mode.



Although the important contact is on the inside, I have marked the contact area in red on the outside surface of the outer plate as well.



If we begin with out two dimensional model in the extended position and flex the parts, we see that the articular edge of the outside plate will travel in an arc until it strikes the upper surface of the outer plate. Again, I have marked that location; this time in blue With the parts of the model disassembled, we can see that the interference is along the articular edge of the outer plate and a line on the side view of the inside plate. This is the critical area of the "secondary lock". Later, I will show that not all "hard" articulations have a secondary lock.



Returning to the three dimensional model, we can see that when the secondary lock is reached, there is contact all along the surface. I have marked up the important edges and surfaces in blue.





Having defined the edges and areas which are important to the articulation, we move on to the areas that may be altered to serve our needs in understanding and constructing armor.

In this diagram, the areas to the left of the blue line on the inside plate are mutable. They may be any shape we need them to be. We may exchange the rectangular box of our two dimensional model or the cylindrical part of our three dimensional model for whatever shape we require. The area between the red edge and the blue band may be made to bulge slightly if needed, but it must stay within the dotted line which represents the travel of the articular edge of the outside plate.

Likewise, the area on the outside plate to the right of the red area can be whatever we would like. The only caveat is that the new shape can not intrude beyond the dotted line which represents the path of the underlapping edge of the inside plate.



This brings us to the point where we can begin to think about applying this articulation to armor.

The diagram which follows shows a variety of simple armor articulations. On the left are plates derived from the generic model by changing the shape of the "mutable zone". The plates on the right are likewise derived by altering the mutable zone of our generic outside plate. The list is my no means exhaustive. Nor are the examples necessarily to scale with one another.

The examples from top to bottom are...
- A sabaton toe cap and the first in a series of lames.
- The ventral plate or "door" of a vambrace lower cannon with an articulating gusset lame at the elbow.
- The distal lame of a gauntlet metacarpal assembly and the first lame proximal to that.
- A gauntlet cuff and wrist plate.
- A greave back with an articulated gusset at the knee.
- A greave front with the proximal lame of an attached sabaton.



Some of these examples speak for themselves, and probably require no further elaboration. Others beg the question of lames in series. I hope to address lame series next.

Mac

[Carter]

In effect, the lock occurs when we try make the longer curve of the inside piece pass through the slightly shorter curve of the outside piece.

Can you elaborate on what makes the inside curve longer then the outside piece?

[Mac]

In effect, the lock occurs when we try make the longer curve of the inside piece pass through the slightly shorter curve of the outside piece.

Can you elaborate on what makes the inside curve longer then the outside piece?

It turns out to be more difficult to measure than I had hoped. Here, I have traced the cardboard pieces of the three dimensional model onto paper so I could clearly mark the paths that are important. I then used a strip of cardboard as a sort of tape measure, and marked the two resulting lengths on that. The lengths are close, but inside path is longer than the outside path. I'm sure the difference would be just a bit greater if it could be measured against the relevant surfaces of the curved cardboard model because of the thicknesses.



Among other things, this points out how small errors in pivot location or wear and tear can make the difference between an articulation locking and popping through.

Than you for keeping me honest here.

Mac

[James Arlen Gillaspie]

Spoiler alert; the effects of synclastic curvatures on the basic principles so far illustrated.

[Mac]

Spoiler alert; the effects of synclastic curvatures on the basic principles so far illustrated.

Indeed. It will get more interesting when I introduce that.

Mac

[Carter]

It turns out to be more difficult to measure than I had hoped. Here, I have traced the cardboard pieces of the three dimensional model onto paper so I could clearly mark the paths that are important. I then used a strip of cardboard as a sort of tape measure, and marked the two resulting lengths on that. The lengths are close, but inside path is longer than the outside path. I'm sure the difference would be just a bit greater if it could be measured against the relevant surfaces of the curved cardboard model because of the thicknesses.

That helps clarify things. Thanks.

[Mac]

At the end of the last installment, I hinted at the idea of lames in the sense that we know them. Today, I'd like to begin exploring that.

If we start with our two dimensional model from before, we can take the articular interface for the outside piece and combine it with the articular interface from the inside piece. These end up "back to back" with some arbitrary length between them. Of course, we get to chose that length just as we get to chose the other parameters discussed earlier.

This diagram is intended to represent that mash up of parts and the resulting lame.



In addition to attaching these lames to more complexly shaped plates, we can string them together. This allows us to add the motions of each individual articulation together, while minimizing the internal protrusion of the underlaps into the space the wearer of the armor hopes to occupy. This stringing together also minimizes the amount of gap which develops between the plates while they are between the extended and flexed conditions. Some people call this type of system "compression articulation". I'm not yet sure if I will adopt that terminology.

The upper figure in this diagram shows a string of four identical lames in the extended position, or the primary lock position. The lower figure shows that same series in the flexed or secondary lock position.



You may be thinking "gauntlets and sabatons" at this point. Hang on to those thoughts. I am going to take a side trip here and come back to them later.

One obvious use of lame series of this type is to combine them in opposite directions around some central plate. In this case the central plate has an underlapping articular interface on each end and a relatively sort space between them. ,

The upper figure in the diagram shows two series of three lames each and a double ended element to join them. The middle figure shows the assembly in the extended position. The bottom figure shows this same assembly in the flexed position.



The idea of a construction like this can lead us to a number of armor application which I will return to in the future. For now, let's consider what it might take to turn this into a "voider of plate" for the back of a knee or the inner bend of an elbow. In such an application, the central plate needs to be narrow, so as not to be too bulky when the limb is flexed. For this, we would need to shrink that space between the pivots. In the most typical voiders, the space between the pivots is reduced completely. That is to say, it's eliminated entirely.

In this diagram, the upper figure shows how we take the double ended underlapping element from before and suppress the entirety of the material from the center of one pivot to the center of the other. If done strictly, it would leave the pivots on rather narrow stems, so we can add a bit of with to keep them strong enough. We can also remove a bit of the material below the pivot.

The middle figure shows how this central element is joined to two series of seven lames. As unlikely as it sounds, the central pivots will go through three thickness of material. I have made the lames narrower than in the previous example so as to be able to fit enough of them in the space allowed.

The bottom figure shows this assembly in the flexed position.



Using similar dimensions, I made up a three dimensional model in cardboard. It's shown here in the extended position. It lays a bit flatter than my two dimensional model did because there is a bit of slop in the pivots. If this were for an elbow, that's what we would require in any case, so it's all for the good.



The same model flexed in hand to give a sense of what the inside looks like.



... and viewed from the side in the flexed position for comparison to the same view of the two dimensional model. Note that the lames are not reaching all the way to a secondary lock position. If we changed the shape of the pieces a bit, these gaps could be reduced, but probably not eliminated. All in all, this assembly has more motion than we need, but that seems to be typical of them.



This last image shows the assembly sitting in the crook of my elbow for scale.



Next time, I plan to return to the idea of double ended assemblies with wider central elements.

Mac

[Indianer]

Mac, I know I'm a little late to the party. I too had trouble wrapping my head around this, but a small change in phrasing worked wonders:

In effect, the lock occurs when we try make the longer curve of the inside piece pass through the slightly shorter curve of the outside piece.

Can you elaborate on what makes the inside curve longer then the outside piece?

Looking at this:

Mac

...I thought...

In effect, the lock occurs when we try make the longer curve of the inside piece pass through the slightly shorter curve of the outside piece.

[The "longer curve is it not the edge length of the inside piece, seeing as the edges of inside and outside piece run parallel and are thus equal in length.
Instead it is the arc that WOULD be drawn along the edge interface across the inner edge were it to pass under the outer edge just a bit past the locking position.]

Doesn't work in the case of the primary lock, as the arc would have to be drawn on a portion of the body of the inside lame that simply does not exist. Instead, it has to be so that the opposite is true: The arc across the contact face on the outside lame is shorter than the arc the inside lame would draw if further extended. But that is very academical altogether...


I also wanna thank You for this - it's a beautiful list of all the places to find the articulation layout. I was amazed at how shortsighted I still am in that regard. Much easier to come up with the "other way around", like on knees and elbows. That may however be just because those are the kind I've mostly thought about until now. Anyway - effectful and astounding pencil work there.

Best Regards
Indi

The examples from top to bottom are...
- A sabaton toe cap and the first in a series of lames.
- The ventral plate or "door" of a vambrace lower cannon with an articulating gusset lame at the elbow.
- The distal lame of a gauntlet metacarpal assembly and the first lame proximal to that.
- A gauntlet cuff and wrist plate.
- A greave back with an articulated gusset at the knee.
- A greave front with the proximal lame of an attached sabaton.



Some of these examples speak for themselves, and probably require no further elaboration. Others beg the question of lames in series. I hope to address lame series next.

Mac

[Mac]

Today, let's start with an assembly of plates and lames. Plates A and B are connected by a series of itentical lames 1, 2, 3, and 4. All the articulations are identical, and proportioned so that when compressed, the upper edge of each one completely covers the previous lame at about the same time it reaches it's secondary lock. The lower figure shows the assembly more or less completely compressed.



Now, let's suppose that we want instead for the whole assembly to be tapered. We want it to be the original height on the left and about 1/2 that height on the right. We can imagine moving each pivot up, along the articular edge until they are uniformly placed above our diagonal line. In effect we are just shortening the length of the three dimensional lame and changing the angle of the bottom edges. All of the lames show an equal "reveal". That is to say the portion which is visible is the same with. This reveal is also parallel sided. It's a handsome looking assembly that satisfy some inner nerdy part of the brain. This is shown in the upper figure.

The lower figure shows what will happen when we compress the assembly as before. Note that while the articulation between plate A and lame 1 behaves exactly as before, the subsequent articulations have increasingly less movement before reaching their secondary locks. If this happens to be enough motion to fill our needs, than we are done planning it out and can move on to making templates.



On the other hand, if we really needed or wanted all the motion that we had in the assembly of uniform height, we need to make some changes. To understand how to plan those changes out, let's take a little side trip. When we consider an individual articulation, we can use a compass to show where the upper edge will travel. By putting the point on the pivot, we can trace out the path of the upper edge of the articular edge until it reaches its secondary lock.



Without changing the compass setting, lets put the point on the primary lock and trace an arc through the pivot.



When we move the point to the secondary lock and trace an arc, it also passes through the pivot. A moment's thought will show that this must be true.



Lets close the compass a bit and make another arc from the primary lock and without changing anything make another from the secondary. These arcs will intersect somewhere above the original pivot location. Now close the compass a bit more and repeat the process. Each of these intersections is a possible pivot location which will produce the same travel of the upper point of the articular edge. Any one of them will work.



When we accumulate enough of these intersection, it becomes clear that they all lie on a straight line. That line is perpendicular to a line between the primary and secondary lock points, and equidistant between them. Note that the variations we see in this picture is all about inaccurate placement of the compass as well as the rather dodgy construction of the same.



Here, I've used the compass to find the midway point and made a bit of correction in the line.



Some of those potential pivot points are not actually on the model of the lames as drawn. Others may be on the lame but do not have enough material to support the pivot. This is of no concern. We can simply redraw the articular edge to include the pivot at the bottom and run that line up the the primary lock point. We also need to adjust the shape of the underlapping edge of the lower plate so give material for the pivot and to provide a uniform underlapp so as to make the primary lock secure.

In the upper figures of this diagram, we can see four possible pivot locations, and the changes to the edges that will be needed for each of them. Note that the arcs which show the travel of the upper point of the articular edge all begin and end at the same points although they differ in radius.

The lower figure shows how we actually chose the locations. In this case, I have drawn a line parallel to the bottom edge to set the spacing of the new pivots from that edge. I have also drawn perpendicular line midway between each of the primary and secondary lock locations that we desire. Where those lines intersect are the new pivot locations. I've adjusted the articular edge lines and the underlap lines as described above. I've also "fudged" it a bit to make everything look even and pleasing to the eye.



The upper figure shows a clean sketch of our newly constructed assembly. The lower one shows what it will look like compressed more or less fully. All of the articulations now flex the same amount, and that amount is equal to our original example. The lame reveals are no longer parallel sided, but the difference is not great enough to be obvious. The inner nerd likes it less, but the inner artist thinks it's an improvement.



Of course, by now you all see where this has been going. The "mutable" part of plate "A" has been changed into a metatarsal plate and that of plate "B" has become a toe cap. The resulting sabaton is similar to the famous 15thC Italian ones in the Royal armouries.






Photo from Nickolas Dupras' thesis.





Photo from Nickolas Dupras' thesis.

The are a couple of different directions that we can take next, so I need to decide what makes the best didactic sense. There may be a slight pause while I think it though.

Mac

[Mac]

Today, I'd like to take a slightly different approach. Let's begin with the idea what we are planning out a Gothic gauntlet. It will have a wrist plate, two matacarpal lames, a metacarpal plate, one cuff lame, and a cuff plate. We will begin by laying it out with all the articulations in their primary lock position, and we will presume that the upper horizon of the entire assembly lays on a more or less straight line.

We can see in the lower figure how this lays on the forearm and hand when the hand is neither flexed nor extended in their anatomical senses. The fingers are shown as if grasping a thing like a sword hilt. The location of the dorsal prominence of the ulnar head is marked with an "x". This important anatomical landmark can not be readily seen when one is looking at his own wrist from this position, but becomes visible and palpable when the hand is fully pronated (as if looking at a wrist watch). I have shown the point of the boss which Gothic gauntlets have to accommodate this prominence. It would be readily seen if the drawing were made from the ulnar (little finger) side, but mostly disappears under the horizon when viewed from the radial (thumb) side.



Designed this way, the gauntlet will have more than enough motion to accommodate the hand in full extension. In many ways, this is "the easy part".



The difficulty arises when the hand is in flexion. The gauntlet I specified above will not bend in this direction. Its primary locks are on a straight line. I thought about drawing it laying foolishly on the back of the flexed wrist, but decided that such a figure wasn't worth sketching.

What I have drawn is the hand in what I believe to be in the position of the greatest flexion that gauntlets normally allow. The wrist will certainly flex farther than this, but gauntlets rarely do.

Above the figure of the forearm and hand, is a curve. This is an approximation where the horizon of an articulated gauntlet must lie. This may seem odd, since the anatomical wrist bends sharply and the forearm and hand remain more or less flat. If we are to simulate that movement with a series of lames, however, we must approximate it over a broader area.



What we need to do is lay out our articulations so that the primary locks allow the lames and plates to fall on that curve. The solution is two fold. The first thing is to let the articular edges "fan out" a bit and the "reveals" of the lames to be greater along the horizon and smaller toward the pivots. The second is to give each lame (and most of the plates) a bit of a curve. This curve is really one of the "secretes" of armor. Armor is almost never really "flat" anywhere, and giving it a bit of "synclastic" curvature not only makes the lames work better, it gives the armor a certain "life" that many reproductions lack.

The upper figure shows the layout of the assembly with the articulations all at their primary locks and the whole thing falling on that curve I showed earlier.

The lower figure shows how the armor will sit on the forearm and hand. Note that the pointy boss on the wrist plate lines up with the dorsal ulnar prominence.



In the next drawing, the upper figure shows the whole of the assembly in the neutral position. The articular edges of the overlaps are more or less parallel and the reveals are more or less even. The underlaps are greater in this position than what we had in the first gauntlet, but this is necessary.

The lower figure shows the assembly resting on the forearm and hand. Note that the locations of the knuckle edge and "gripping notch" remain in the same place with respect to the hand. The pointy boss on the wrist plate is also over the ulnar prominence.



The upper figure in this illustration shows the theoretical limits of how much motion the assembly is capable of before reaching the secondary locks.

The lower figure shows the practical limits. Note that while the important locations on the metacarpal are still where they must be, the wrist plate have slid back on the arm slightly. While this moves the pointy boss to a position distal to the ulnar prominence, it is not a problem because the plate can not bear on the prominence when the wrist is extended.



Two important elements have been introduced here. The first is setting the articular surfaces of opposite ends of lames and plates at an angle to one another when in the primary lock. This "fanning out" of lames will be critical to articulations I plan to describe in the future. The other important element is synclastic curvature between the articular surfaces. This not only makes the armor look smooth, it helps to generate internal clearances for underlapping edges while at the same time reducing the gaps when the articulation is in between its primary and secondary locks. More on this later.

Mac

[Sean M]

I did not realize that the plate with the ulnar bump in Gothic gauntlets needed to be dished with the hollow part down! More fool me, I need to stare at more pictures of these fantastical futuristic devices.

[Mac]

I did not realize that the plate with the ulnar bump in Gothic gauntlets needed to be dished with the hollow part down! More fool me, I need to stare at more pictures of these fantastical futuristic devices.

It's a bit more complex that that, really. These can be made to work without the general synclastic curvature by making the articulations much sloppier and/or turning the underlapping edges down more or less locally. The effect is a bit "choppy" looking. The nicer examples have an overall curvature.

Here is one with quite a curve.



Now... that said, I have probably overstated the curves in my sketches. Perhaps I will go back and tone them down a bit an re-scan them.

Mac

[Batteredshell]

I think these drawings/diagrams are absolutely fantastic! Can't wait to see you keep going!

[jdr1066]

Mac, this is the greatest thing since... your last blog post. I am checking in every day for the next installment...

Thank you.

John Rucker

[Mac]

Mac, this is the greatest thing since... your last blog post. I am checking in every day for the next installment...

Thank you.

John Rucker

I think these drawings/diagrams are absolutely fantastic! Can't wait to see you keep going!

Thank you, gentlemen!

I'm working on the next post, and hope to have something tomorrow.

Mac

[Ian BB]

I just wanted to say how grateful I am you're putting all this down on paper Mac. I often read the forum and I don't comment but I wanted to make sure it was said.

[Mac]

I just wanted to say how grateful I am you're putting all this down on paper Mac. I often read the forum and I don't comment but I wanted to make sure it was said.

Thank you!

Mac

[Sean M]

Now... that said, I have probably overstated the curves in my sketches. Perhaps I will go back and tone them down a bit an re-scan them.

Mac

I think I can see a bit of the doming in Wade's photos of his Gothic gauntlets. I should get out a Windows box and look at the Iron Key.

[Mac]

In the last installment, I showed how we can plan out an articulated assembly by letting the parts fall on a curve, rather than a straight line. Today I'd like to take that further by showing a series of identical or near identical lames that all fall on a relatively small radius.

This diagram shows an assembly of four identical lames. In the upper figure, they are all in their primary lock positions and fall on a portion of a circle.

In the lower figure that same assembly has been straightened out till the lames are in their secondary locks.



There are two basic ways to make mitten gauntlets. They can be made to approximate the movements of the fingers as a smooth and continuous curve or they can be made to bend at each of the three places where fingers bend. I will address the smooth type today and deal with the angular type in a future post.

The assemblies I have just shown have uniformly shaped lames. This is not our end goal today, but I think we should start here because they are easier to understand.

Lets start with a "generating circle" of an arbitrary radius. For this example I have used 1 3/4" or 44mm. I have drawn a vertical and horizontal at 90°, and with the compass at the same setting, I've found the locations of the lines which bisect those at 45°. These lines were chosen for ease of layout.



The inner circle represents where the pivot points will be, and this has arbitrarily be set to make the space between the two circles be 5/8" or 16mm. Where the inner circle intersects the layout spokes I've drawn circles of about 3/8" or 10m diameter. These represent the amount of metal that will support the pivots.



In this picture I am using a machinist's angle gauge set at 45° to scribe out the locations of the "lock lines". The vertices of the angles are set at the places where the spokes cross the generating circle. Note that the angle between the lock lines is is equal to the angle between the spokes. This will yield articulation which will move the amount we need to go from the circle to straight.



In this picture, I've drawn lines parallel to the left hand lock lines and tangent to the pivot circles. This will form the underlapping material of the lame. In addition to that I've drawn lines to represent the lower edges of each lame. These are tangent to adjacent pivot circles.



Here, the underlapps are shown dotted and the articular edge of the lames are darkened. Some of the construction lines have been erased just to emphasize the edge lines.



Lastly, the upper curved horizon lines are established. Note how they follow the construction circle until just before the lock lines on the right. The horizon lines then dip down inside the construction circle by about one metal thickness and continue that way till they reach the other lock line which marks the edge of the underlap.



This will produce the four lame model at the beginning of the post. If one manipulates the elements of the model on the light table or in bits of flat cardboard, the articulations will go from one lock to the other freely. Unfortunately one can not conveniently model this in three dimensional cardboard because it relies on the synclastic curvature of the plates to make the locking surfaces be in the right place, as well as providing clearance for the underlapping edge.

As I said at the beginning of this post, we will need something a bit different from the first example to make a proper mitten gauntlet finger assembly. This is because fingers taper toward the tips, and the depth of the lames should taper as well. A series of identical lames would work, but a series of lames that tapers distally works better, looks better, and is by far the more common thing in authentic work.

Much of the design of a model for the tapered assembly will be similar to what I have shown above, but with a couple of differences. I will also begin with dimensions which have worked for me in the past. The resulting model with feature five lames in total. That seems to be the most common number for full length finger assemblies. The next most common is seven. I don't know why, but it seems to be so.

Like before, we begin with a generating circle. In this case the radius will be 1 5/8" or 41mm. This should make a good "size medium".



Next, we set the compass to the distance we want between pivots. I have found that 11/16" 16.5mm works well for the radial (thumb) side of size "medium" gauntlets.



With the previous example we set all the pivots on one circle. Here, because the lames will get shallower as we move out to the finger tips, we are going to use a series of radii. Each successive pivot will be closer to the generating circle by about a metal thickness. For the first one, we will set the compass to strike an arc that's 9/16" 14mm below the generating circle. The place where the two arcs intersect is our first pivot location.



We then set the compass at that pivot and strike another arc at our pivot spacing distance (11/16" 16.5mm). The compass is then returned to the center and set for the next pivot radius, which we recall is about a metal thickness closer to the generating circle. This process is continued until we have set six pivot locations. The last one will not actually be a pivot, but only serves to make the terminal lame be a similar size to the others.



The next place where the procedure differs from the previous one is that each pivot will need a different angle between the "lock lines". This is because the angles between the "spokes" are not the same, but get smaller from proximal to distal in the series. So.... we use the angle gauge to measure the angle between the first two spokes.....



... and transfer that to the first pivot to draw the lock lines. We can just leave the angle gauge set, and not even look at the number. The gauge is set so that spoke evenly bisects the angle and the lines are traced against the gauge.



It should look like this.



Here we see that the angle between spokes 2 and 3 is a bit smaller than that between 1 and 2. This is as it should be.



Adjust the angle gauge to match this angle and transfer the angle to the location of the second pivot.



We continue measuring each successive spoke angle and transferring that angle to the next pivot till we run out.



From here on in it's more like the first example. This shows how a 3/8" 10mm circle is made around each pivot point, and also the non-pivot point at the end of the series.



Like before, lines are drawn parallel to the left hand lock lines and tangent to the pivot circles.



The lower edges of the lames are established with lines tangent to adjacent pivot circles.



Here we see all the important lines darkened. Like before, the upper horizon line deviates from the generating circle just before the right lock line to allow for underlap. In this case I've shown the terminal lame finished off with a sunk border and roping. Any part of that lame past the right lock line can be altered and it will still work. I've also shown the angles what were generated for each articulation. These angles will be different if any of the other parameters are changed, but they will always get smaller from proximal to distal.



This last image shows how the assembly will interact with the fingers flexed and extended. The thumb has been omitted for clarity.



In a future post, I will address what happens at the proximal knuckles.

This has been surprisingly difficult to write up. I hope that you will point out anything that is not clear so I can edit to clarify.

Mac

[Carter]

Next, we set the compass to the distance we want between pivots. I have found that 11/16" 16.5mm works well for the radial (thumb) side of size "medium" gauntlets.

Is the distance between pivots closer on the ulnar (pinky) side? What effect, if any, would the pivot axes not being parrel to each other have on the articulation / planning?

[wcallen]

Mac,

That is very well thought out and described.
Thank you so much for putting all the work in to show the theory behind these things.
I own one piece that reflects a (well done) version of what you are describing in the latest post. I had to take a new picture to show it properly curled up to reflect your drawings. It is annoyingly troublesome to get it to stand up by itself in this pose. Anyway, here it is:


I think that your work models this style of articulation well.

Wade

[jdr1066]

I have been doing this by sight and instinct for all these years, and all along, there have been people who actually understand it...

Again, Thank you!

[wcallen]

and... the same gauntlet fingers straight:


Wade

[Mac]

Next, we set the compass to the distance we want between pivots. I have found that 11/16" 16.5mm works well for the radial (thumb) side of size "medium" gauntlets.

Is the distance between pivots closer on the ulnar (pinky) side? What effect, if any, would the pivot axes not being parrel to each other have on the articulation / planning?

This is an excellent question. Yes. The pivots are closer on the ulnar (little finger) side. I did the modeling from the radial (thumb) side just so I could show it overlaid on a hand sketch.

In practice the difference how the two sides behave is quite small, and a model for the one side will probably work OK for the other. It might be "best practice" to model both sides. I suspect that a good compromise between what I showed and best practice might be to model the pivot spacing as a distance in between the two. As I say, though, the difference between that and what I showed will be quite small.

Mac

[Mac]

As an addendum to the previous installment; here are a couple of template sets.

This first one is from a medium sized Maximilian gauntlet. I've included a ruler for scale. Note that the pivot spacing is always a bit shorter on the ulnar (little finger) side. There are some notes about material I trimmed away during assembly etc.



This set is from a size medium gauntlet with a "zig zag" finger profile. The notches represent the "ridges and valleys", which I marked on the flat stock before cutting out the pieces.



The second set has been used many times, and is part of one of my standard "sportified" hourglass gauntlets.



Mac

[sirfelix]

thank you once again Mac!
what are the chances of posting the rest of the sportified" hourglass gauntlets? mostly the thumb and metacarpal
thank you.

[Mac]

thank you once again Mac!
what are the chances of posting the rest of the sportified" hourglass gauntlets? mostly the thumb and metacarpal
thank you.

I was going to do that yesterday, but I noticed that this set does not have a cuff template. I've just searched the likely places and haven't found it. I'll post a set for the larger size in a bit. They won't just go on the scanner bed, and I need set up the camera and a background.

Mac