275 lines
13 KiB
HTML

<html>
<head>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-129280539-1"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag() { dataLayer.push(arguments); }
gtag('js', new Date());
gtag('config', 'UA-129280539-1');
</script>
<title>Bellows</title>
<meta content="True" name="vs_showGrid">
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<link rel="stylesheet" type="text/css" href="/Site.css">
<script type="text/javascript" src="/menu.js"></script>
</head>
<body>
<div id="mainContainer">
<div id="topBar">
<h1 align="center">Bellows</h1>
</div>
<div id="leftColumn" w3-include-html="/menu.html"></div>
<div id="mainContent">
<p>The linear slides I got off ebay didn't come with covers to stop swarf getting to the rails and ballscrews. So I decided to cover them with a bellows.</p>
<p><img border="0" src="bellows.jpg" width="476" height="377"></p>
<p>I did a little bit of research. There are some stock sized
commercial bellows available; these are usually rubberised fabric, welded at the seams.&nbsp;
I couldn't find any that fit, nor could I afford them.&nbsp; So
I googled for DIY bellows folding.&nbsp; Again, I drew a blank.&nbsp;
Then I checked the USPTO, and I hit a gold.</p>
<p>
<a href="http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/PTO/srchnum.htm&r=1&f=G&l=50&s1=6,054,194.PN.&OS=PN/6,054,194&RS=PN/6,054,194">Patent number 6,054,194</a>,
invented by Nathan R. Kane, is not just a description of
bellows, but it is a complete thesis on optimal fold patterns
for bellows.&nbsp; The optimal patterns will maximise the
bellows extension length, while minimising the side wall
distortion.&nbsp; By using just the fold pattern, a bellows can
be created that holds its shape without external support.</p>
<p>Polypropylene sheet was selected to make the bellows.&nbsp;
Polypropylene is a good &quot;hinge&quot; material, meaning it can be fold
back and forward repeatedly with fatigue.&nbsp; I tried a couple
different thicknesses, 1.2mm, 1mm, .7mm, and finally settled on
0.39mm - not that this is the best size, its just the thinnest
stuff I could find in large sheets.&nbsp; I think thinner sheet
would be better, but I couldn't find any.&nbsp; I found it at my
local craft store.</p>
<table id=polypropyleneGrain margin=5 bgcolor="#FFAAAA" border="2" bordercolor="#FF0000" cellspacing="6" cellpadding="6" bordercolorlight="#FF0000" bordercolordark="#FF0000" style="border-collapse: collapse">
<tr>
<td>
<p>Update!</p>
<p>Just a note on polypropylene.&nbsp; Polypropylene has grain!&nbsp;
I'm sure my terminology is incorrect, but polypropylene, like
paper or timber, has a preferred strong direction and a weak
direction.&nbsp; If you try to tear polypropylene sheet in its
weak direction, it will continue to tear in the direction it is
forced.&nbsp; In the other direction, 90&deg; to the weak
direction, the polypropylene will not tear, in fact it will turn
90&deg; to the weak direction.&nbsp; This is because of the
polymer chains (so I've read).&nbsp; </p>
<p>Why is this a problem?&nbsp; If you try to fold the
polypropylene with the folds running parallel to the weak
direction, the polypropylene will crack.&nbsp; It is important
to get the direction correct so the folds are perpendicular to
the weak direction.</p>
<p>So how do you find the weak direction?&nbsp; On the corner of
a sheet of polypropylene, make a small cut, 5mm long and about
5mm from the edge.&nbsp; Grab this tab and pull it.&nbsp; The
pictures below show the results.</p>
<p>
<table>
<tr>
<td><img border="0" src="TearAgainst.jpg" width="250" height="356"></td>
<td><img border="0" src="TearWith.jpg" width="250" height="356"></td>
</tr>
<tr>
<td align=center>Against the grain</td>
<td align=center>With the grain</td>
</tr>
</table>
<p>
The photo on the left shows trying to tear against the grain -
you can't do it.&nbsp; The photo on the right shows tearing with
the grain.&nbsp; The polypropylene tears easily.&nbsp; It is
imporant that the bellows fold lines are placed perpendicular to
this weak direction.</p>
</td></tr></table>
<p>&nbsp;</p>
<h2>Building the Bellows</h2>
<h3>1) Print the fold pattern</h3>
<p>The first step is to produce a fold pattern for the bellows.&nbsp;
For my first bellows, I used MS-Visio to lay out the lines.&nbsp;
This was a bit tedious, so I wrote a simple program.</p>
<p>
<img border="0" src="Application.png" width="669" height="559"></p>
<p>The program will only create a fold pattern for a simple bellows with either
regular or alternating folds.&nbsp; The &quot;generate g-code&quot; option
is not functional.&nbsp; I want to be able to generate g-code to
guide a knife to score the bellows material, however all the
bellows I need to create are larger than my mill, so this
feature was postponed.</p>
<p>The parameters are...</p>
<table id="table1">
<tr>
<td width=150px>Config</td>
<td>Different configurations can be selected.&nbsp;
Configurations are automatically saved when the program
exits, or a new configuration is selected.</td>
</tr>
<tr>
<td>New</td>
<td>Creates a new configuration.&nbsp; </td>
</tr>
<tr>
<td>Shape</td>
<td>Only a &quot;Half Cover&quot; is supported (2 sides and a top)</td>
</tr>
<tr>
<td>Inversion</td>
<td>Only 2 inversion are supported</td>
</tr>
<tr>
<td>Mounting Folds</td>
<td>The total number of extra folds to add for mounting
the ends of the bellows</td>
</tr>
<tr>
<td>Alternate Folds</td>
<td>If true, adjacent folds will alternate up and down,
rather than being in the same direction.&nbsp; This
makes the compressed bellows smaller, but the top of the
bellows has folds that may collect swarf. </td>
</tr>
<tr>
<td>Inside Width</td>
<td>The inside width that the bellows is going to cover.</td>
</tr>
<tr>
<td>Inside Heigth</td>
<td>The inside height that the bellows is going to
cover.</td>
</tr>
<tr>
<td>Protected Length</td>
<td>The length of the area the bellows is going to
cover.</td>
</tr>
<tr>
<td>Fold Width</td>
<td>The width of one fold.</td>
</tr>
</table>
<p>The program will print out the fold pattern.&nbsp; It will
tile the output over multiple pages.&nbsp; The diagonal line is
used for lining up the pages when they are stuck together.</p>
<p>The image below shows a snippet of the fold pattern.&nbsp;
Note the solid and dotted lines.&nbsp; A solid line is a peak
fold, and a dotted line is a valley fold (or vice-versa, it
doesn't matter).</p>
<p><img border="0" src="Pattern.png" width="451" height="447"></p>
<p>The picture shows the fold pattern with a sheet of
clear/translucent polypropylene on top.</p>
<p><img border="0" src="Layout.jpg" width="600" height="435"></p>
<p>The program and source can be downloaded here
<a href="Bellows.zip">
<img style="background-color=white" border="0" src="../../gplv3-88x31.png" width="88" height="31"></a>.&nbsp;
The code isn't terribly stable.&nbsp; It was only used to
generate a couple of patterns.&nbsp; The displayed pattern will
only be updated when you tab between fields.&nbsp; There is no
zoom - make the window bigger to see more of the pattern.</p>
<h3>2) Trim</h3>
<p>The polypropylene sheet is cut to size.&nbsp; It is then
stuck to the fold pattern using tape.</p>
<h3>3) Score the lines</h3>
<p>The next step is to encourage the polypropylene sheet to fold
at the correct place.&nbsp; If these were being produced
commercially, they could be vacuum formed, or pressed.&nbsp;
Since this is overkill for a couple of one-offs, I tried a
couple of manual techniques.</p>
<p>First I tried scoring with a hobby knife.&nbsp; This was
fine, but in a few spots the cuts went a bit deep and made the
fold weak.&nbsp; Next, I tried using my soldering iron to melt a
grove.&nbsp; This wasn't terribly successful as shown by the
video below.&nbsp; I did learn though that the scoring and
folding needs to be very accurate, or the bellows will not fold
properly.&nbsp; Finally, I settled on a scriber - a pen with a
sharp carbide tip used for scratching metal, to scratch the
polypropylene.</p>
<p>Use a metal ruler and carefully trace over the peak folds.&nbsp;
Take care and do this accurately or the bellows will not fold
properly.&nbsp; Don't score the diagonal guide line running the
length of page.</p>
<p>The picture below shows the scored peak folds.&nbsp; The
lighting in the picture makes them appear black.&nbsp; </p>
<p><img border="0" src="ScoredTop.jpg" width="500" height="362"></p>
<p>When scoring fold corners (line intersections), make sure you
overshoot the intersection by 2 or 3mm.&nbsp; This will help
with the folding later.</p>
<p>
<img border="0" src="ScoredOver.jpg" width="682" height="511"></p>
<p>When the top is finished, remove the polypropylene sheet,
flip it over and stick it back down on the fold pattern.&nbsp;
You can use the scored peak folds to line up with the printed
peak lines as the bellows patterns are symetrical.&nbsp; Then
score the valley fold lines.</p>
<p>The photo below shows a scored polypropylene sheet.</p>
<p><img border="0" src="Score.jpg" width="682" height="511"></p>
<h3>4) Crease the Folds</h3>
<p>Next, each of the scored lines needs to be creased.&nbsp;
Here, we are not trying to fold the bellows, just get the folds
going in the right direction.&nbsp; Just bend each fold and
squeeze them with your fingers, like in the photo below.</p>
<p><img border="0" src="Crease.jpg" width="500" height="303"></p>
<p>It is important to get all the small angle folds.&nbsp;&nbsp;&nbsp;
The folds must also be done in the right direction.&nbsp; If a
scored line is on the top of the sheet, the two sides of the
fold should fold down, like this...</p>
<p>
<img border="0" src="FoldDirection.png" width="497" height="312"></p>
<p>The creased bellows will look like this...</p>
<p><img border="0" src="Creased.jpg" width="568" height="407"></p>
<h3>5) Folding</h3>
<p>The final folding is the tricky bit.&nbsp; Although the
initial creasing does help the bellows want to go in the right
direction, it still takes a lot of finger muscle and patience to
fold the bellows.&nbsp; It is important that the corners are
sharp, otherwise they will be a source of ballooning.</p>
<p>I found it easiest to fold the bellows if I worked on one
side, folded 3 or 4 folds, clamp it, then do the other side.</p>
<p>
<img border="0" src="FingerFold.jpg" width="568" height="426"></p>
<p>The quick release clamps worked well holding a few folds at a
time.</p>
<p><img border="0" src="AndClamp.jpg" width="568" height="358"></p>
<p>Finally, all folded.</p>
<p>
<img border="0" src="DoubleClamp.jpg" width="594" height="248"></p>
<p>The bellows is clamped between two chunks of timber to
encourage it to stay in place.</p>
<p><img border="0" src="BigClamp.jpg" width="409" height="408"></p>
<p>The completed (blurry) bellows.</p>
<p><img border="0" src="Folded.jpg" width="392" height="352"></p>
<h3>6) Mounting</h3>
<p>Obviously this will depend on what you plan to use the
bellows for.&nbsp; I used some 3mm steel plate to mount on each
end of the bellows.&nbsp; The bellows are held to the plates by
M4 screws and small lengths of 3x10mm bar, tapped for the M4
screws.&nbsp;&nbsp; These were then mounted to my X axis slide.</p>
<p>
<img border="0" src="MountingParts.jpg" width="417" height="362"><img border="0" src="Mounted.jpg" width="414" height="343"></p>
<p>3mm steel plate probably wasn't the best choice for mounting
the bellows, as I have already hit tall hold down bolts with it.&nbsp;
In the future, I may tray thick plastic, and self adhesive
velcro strips.</p>
<h2>When good bellows go bad!</h2>
<p>The video below shows what happens when a bellows isn't
folded correctly.&nbsp; The bellows on the right was bubbling
out when it was closing.&nbsp; This snagged when moving in the Y
direction, leaving what you see now...</p>
<object width="425" height="350"> <param name="movie" value="http://www.youtube.com/v/1O0ln1kLFCI"> </param> <embed src="http://www.youtube.com/v/1O0ln1kLFCI" type="application/x-shockwave-flash" width="425" height="350"> </embed> </object>
</div>
</div>
<!-- #include file="/analytics.html" -->
</body>
</html>