Welcome to the Podiatry Arena forums

You are currently viewing our podiatry forum as a guest which gives you limited access to view all podiatry discussions and access our other features. By joining our free global community of Podiatrists and other interested foot health care professionals you will have access to post podiatry topics (answer and ask questions), communicate privately with other members, upload content, view attachments, receive a weekly email update of new discussions, access other special features. Registered users do not get displayed the advertisements in posted messages. Registration is fast, simple and absolutely free so please, join our global Podiatry community today!

  1. Everything that you are ever going to want to know about running shoes: Running Shoes Boot Camp Online, for taking it to the next level? See here for more.
    Dismiss Notice
  2. Have you considered the Critical Thinking and Skeptical Boot Camp, for taking it to the next level? See here for more.
    Dismiss Notice
  3. Have you considered the Clinical Biomechanics Boot Camp Online, for taking it to the next level? See here for more.
    Dismiss Notice
Dismiss Notice
Have you considered the Clinical Biomechanics Boot Camp Online, for taking it to the next level? See here for more.
Dismiss Notice
Have you liked us on Facebook to get our updates? Please do. Click here for our Facebook page.
Dismiss Notice
Do you get the weekly newsletter that Podiatry Arena sends out to update everybody? If not, click here to organise this.

Vacuum formed versus direct milled orthotics

Discussion in 'Biomechanics, Sports and Foot orthoses' started by Simon Spooner, Sep 1, 2010.


Which orthotic manufacturing process do you prefer?

  1. vacuum formed

    12 vote(s)
  2. direct milled

    11 vote(s)
  3. no preference

    4 vote(s)

  1. Members do not see these Ads. Sign Up.
    I should be interested to know how many people prefer direct milled devices over vacuum formed devices? Further whether there are specific features of devices manufactured using these techniques which provides performance benefits over one another?

    If you use vacuum formed why?
    If you use direct milled why?
    If you don't care either way, why not?
  2. Graham

    Graham RIP

    I Have used both.

    I have generally found that DM devices do not have the flexibility of the molded devices. I use XT and now performance Rx ( a nylon based shell material). The DMs were tolerated reasonably well but I do not use these any more as the molds are tolerated much better, with better clinical results, even with over weight clients.

    I use these in a Rx based on Sagittal facilitation principals.
  3. Thanks for your response, Graham. Why did the direct milled devices not have the same 'flexibility"? Was it simply that they were milled out of a different material? The stiffness of the device should be defined by the geometry and material, so if we had two devices of identical geometry, one vacuum formed and the other milled, they should have the same stiffness characteristics- right? I guess with vacuum forming you start off with a sheet of pretty uniform thickness plastic, but as it is pulled over the model it will stretch to a greater extent over more curved features? Does this pulling and heating also "pre-stress" the plastic and/or change the material properties?
  4. Graham

    Graham RIP

    Very possibly! I found the DM devices to be very stiff through the arch and my XT and Nylon devices to have a greater "spring" like tendency. However, I do not use poly pro molded as I find these too "stiff" so perhaps it is just the choice of material that makes the difference.
  5. Does anyone know of the proportion of labs offering only direct milled devices versus only vacuum formed devices? Is the traditional vacuum formed manufacturing process falling out of favour and being replaced by direct milling? Are there any labs that once offered vacuum forming, but now only offer direct milling? Perhaps more telling: are there any labs that started up as direct milled only, but now offer vacuum forming?

    One last question: is there a directory of commercial labs?
  6. joejared

    joejared Active Member

    Of the labs I've worked with that have been around more than 1 year in my network, most no longer manufacter positive molds. Less than 5% are positives. Less than 25% actually have plaster based facilities.

    Vacuum forming is here to stay, at least for the diabetic line of products and some labs, although I recently added the capability to manufacture a full length soft accomodative device. The reasoning behind his opinion is that often, the top cover is glued to the material prior to forming the device, which is much easier than putting the top cover onto the device afterwards.

    For hard shell devices, use of positive molds is sparse, limited to styles such as UCBL.

    You could check out lower extremity review's resource guide for starters as well as my own customer list.
    Last edited: Sep 1, 2010
  7. Graham

    Graham RIP

  8. Thanks Joe.

    Here's another question I have: when orthoses are direct milled what proportion of the original block of plastic ends up as waste and what proportion ends up as product?

    Here's an interesting study: weigh the block of plastic pre-production, weigh the devices produced from the block.
  9. Thanks Graham, I've just started using a commercial lab again as I simply don't have the time to do all my own stuff anymore: I've gone for the Langer Group UK http://www.lbguk.co.uk/ as they too can provide both direct milled and vacuum formed- so far so good, very happy with the devices to date, just trying to think toward the future a little and whether I should be trying more direct milled stuff.
  10. joejared

    joejared Active Member

    Identical thickness isn't necessarily identical. When you measure a device, you're measuring at the top side of a radiused groove to the opposing top side of the radius. If you take this into account by volume, they should be closer. Typically, at least in OreTek, the typical machining step-over is 1.5mm apart using a 1/2" ball nosed endmill. A 1mm stepover would have considerably less ridging. In Autocad, I tried to paint a better picture of this, which is in inches. For a 2mm stepover, the actual difference in thickness from valley to valley is 0.6mm counting both sides of the device. With a stepover of 1mm, the difference is 0.15mm The drawing is in inches, and both doubling and conversion has already been done. To match thicknesses, you'd have to take this factor into account.

    Last edited: Sep 1, 2010
  11. So, if I'm reading you correctly Joe, basically because of the tooling in the direct milled devices rather than having a relatively uniform surface shape like we obtain via vacuum forming, the milling machine path creates a series of "furrows" and ribs. Obviously the direction and depth of these "furrows" will influence the stiffness of the device. I guess most manufacturers will use an optimal tool path to minimise manufacturing time, but should we be looking at using the tool-path as part of the prescription, purposely defining the direction and depth of the furrows / ribs? Interesting. One of the things that excites me about direct milling is the opportunity to not have uniform thickness shells.
  12. joejared

    joejared Active Member

    The ribs have some effect with thicknesses are marginal, but it's the furrow that has majority effect. My own patented method creates a ribbing that would naturally strengthen the device, but it's really only a time saver, causing a constant velocity during manufacturing. (See videos at the top of this page) Machine tolerances are much more important than the ribbing effect.

    This thread also partially explains why my theoretical zigzag gauge cut position is left of where it should be in establishing final plastic thickness tolerances. Gauge cut is something done on the machine to find the top surface of the plate and is rarely used for anything other than machine setup.

    Further. it argues at least for plastic thickness that I need to add a suggested design code based on wrk# variables to take this into account and complete my equation parsing routines to include trigonometric functions, although a simple y=mx+b style of code will likely be close enough.
    Last edited: Sep 1, 2010
  13. A straight line, you lost me. Do you use different direction in the furrows in different areas of the device? Do you purposely use wider / deeper furrows orientated at specific angles in different areas of the superficial surfaces of the devices?
  14. joejared

    joejared Active Member

    Yes. It's all covered in patent #6,865,442, but I'll send you a private link to my drawings page so you can see it in 3D. In answer to your other questions, OreTek has the ability to control heel thickness, lateral thickness, overall thickness, as well as the thickness along the edges, commonly referred to as feathering. For valgus posted devices, it's not uncommon to thicken up the device laterally, whereas for sport, pump, dress styles, the inverse is more common. I don't worry about the grooves, but rather, keep machine tolerance as a focus so that what's asked for is delivered, typically with a goal of +-0.15mm and in practice, +-0.3mm.
    Last edited: Sep 1, 2010
  15. robby

    robby Active Member

    one issue that no-one has mentioned so far is the ability of DM being able (assuming that the company being used has the software) to mill different areas of one device to different levels of thickness giving totally contralable flexibility in different areas of the shell. This is the bit that I find most useful. I can have a flexible arch profile with a solid lateral border or solid heelcup, or a very solid arch with thiness elsewhere in the shell. I have total control over this with milling, which you cannot have with vacuum pressed devices, the thickness is the thickness and the property of material is maintained.

    also using true material properties, we should remember that with vacuum forming you are heating the plastic /nylon for a second time (once in manufacturing) and this will and does cause changes in the bonding in the chemical structure of the materialm and can change the flexural properties of the shell, this does not happen with milling giving a far more stable material.

    regarding materials available I use Polyprop, Polyeth, and Eva all milled which when used in combination with scanning (especially for TCIs) give the best fit I have ever had from commercially made devices.

    Like Simon, I use Langer now, (have always used them in the past for Vacuum pressed and still get them too). I used to use Talar Made Custom, which Langer took over last year and with the improvements thay have made i now use the Langer group exclusively for my orthoses.

    They still use POP positives where appropriate, and with RX labs (also n ow owned by The langer Group) always use POP.

    I have used other labs at times, but found that they cannot give me the full range I need and use. I like the one stop shop I get there.

    Joe perhaps you should give Langer UK a try, if you want POP using! Langer US no longer owned by Langer now under the wing of TOG, what would Sheldon Langer think????

    I use different types and production methods to suit my patient needs, DM for sports devices as I can get well fitting orthoses with good range of felxibilities and vacuum pressed for other uses such as in city/office wear. I also use RX lab (Rootian ) devices where needed.

    the basis and most important thing to take note of when using DM devices is the software of the company, this is by far the most important bit of the whole production process, any mill will do as it is told by the driver, but the driver needs the correct software to make good orthoses it needs to be a system that design TRUE custom devcies and not uses library shapes morphed to fit.


  16. Said it here:
    You see that's where I get confused, why can't we create all the devices we need using the same production method? Why do you perceive a need to use different production methods for different devices?
  17. joejared

    joejared Active Member

    Actually, I just did. ;-) Additionally, one customer asked for and was given an update to create a rib under the arch, very much localized like a fascial groove. The majority of the device can now be thinner for that design as a result, simply because it has reinforcement where needed.

    This argues for a possible conveyor based tempering process that passes through an oven to temper the materials to match, or alternatively, flick them into an oven for a period of time after machining. Of course the oven would need to be set to a temperature that would be appropriate for tempering of the material that is machined.

    Um, why?

    Both software and hardware are equally important. Scanners, routers/machining centers need to be able to emulate correctly actual casting. So does the software and the product is only as good as the weakest link.
  18. joejared

    joejared Active Member

    We can. Matching flexibility to a hard shell device isn't rocket science. In fact, it's barely 7th grade math. More recently, my newer customers have been creating rigidity codes with only minor variances based on arch height. They enter 0.00 for material thickness and specify overall rigidity in a design code, and forget about having to do the math. A previous post, covering wrk# variables shows one such example. The only variation from this theme, an extremely high arch is stronger because thickness is measured vertically, from top to bottom, and not tangent to the surface.
  19. Joe, I appreciate your responses, but you need to spell out your acronyms, what's a wrkhash variable is this something that applies only to your manufacturing process?

    How are you measuring orthotic rigidity?
    P.S. don't take this the wrong way, Joe as I'm a huge fan, but is it just me or does Joe's new avatar look like a young Brian Eno circa Roxy Music?
  20. ENO!:drinks (glass in hand- good effort, sir)

    Attached Files:

    • eno.jpg
      File size:
      287.5 KB
  21. joejared

    joejared Active Member

    wrk# is a series of design code variables. I linked it in a previous message but here is the body of the link:
    Variable Name: wrk0 ~ wrk10
    Variable Name: wrk0 thru wrk10
    Type: numeric
    Purpose:temporary storage
    Detailed Description:Temporary storage work variables

    Wrk variables are temporary storage variables, intended to be an aid in obtaining a more complex result than a single equation can give.
    In the example below, a material thickness is calculated with a boundary of 2 to 6mm. 2 weights are listed, 100 lbs and 200, and their 2 thicknesses, 2mm and 4mm, respectively. In the resulting equation, y=mx+b (point slope), plastic thickness is determined.
    pt=pt > 2
    ;wrk0 calculates slope
    ;wrk1 calculates Y intercept
    ;y = thickness x = weight
    pt=pt < 6
    I only make orthoses for research and samples for potential customers, but actual measurement would likely involve some sort of force measurement deflecting the surface a given distance or alternatively, a subjective test.
    In practice, laboratories define thickness based on rigidity. wrk# codes just simplify their calculations.
  22. joejared

    joejared Active Member

    Only when I wear lipstick.

    And yea, I make being 45 1/2 look good. :)
  23. robby

    robby Active Member

    we can do that, milling systems can mill different properties into the material, but we are limited to the materials available for example I cant get milled Toprelle- a rubber based material that I use for some devices in older less flexible patients.

    you cant mill carbon fibre or TL2100 for example. so I use Vacuum forming for these.

    Software and hardware are important BUT once you get to a certain level of milling machines etc then the software becomes more important, and when working on an industrial scale most companies have the correct hardware BUT very different software, In fact on some factory/lab visits I have made I found that companies had EXACTLY the same hardware but different software which made massive difference to the finished orthoses .
  24. Phil Wells

    Phil Wells Active Member


    May be worth introducing the commercial variables to this discussion.
    It can take as little as 30 minutes to scan, design, machine and polish a direct milled device.
    I don't think vacuum forming can match that.

    Also I would question Robby - is that Rob Bradbury Sales and Marketing Manager of Langer - as this is his previous pseudonym - as it did seem like a bit of a sales pitch. If not I apologise in advance.

    Also all the hype about calculating methodology for material thickness is a bit much. Without an understanding of the actual GRF of the individual patient, within the actual shoe etc, etc , etc it all becomes a moot point to the clinical outcome.

    Re vacuum forming, it is worth looking at AFO manufacture to get an idea of the variability's available. When forming the polyprop over the cast, the technician (as per the Rx) can thin the plastic by stretching to effect certain flexibility changes. It is highly possible that this can happen when pressing the FFO shell - not sure how significant this is if it occurs at all.
    I remember from my time at Langer that the shells most commonly fractured at the point where the rearfoot post ended. On examination the polyprop was thinner in this area - an artefact of the vac process or 'necking' - I don't know. With direct milling you can thicken that area easily to prevent this from happening.

    Re the question about waste, I have estimated that with direct milling it is approx 75% but with a little bit of forward thinking, this can be recycled.
    Direct milling also offers new materials not available with vac forming. E.g LD Polyurethane's - we are finding that these last approx 5 times longer than the equivalent EVA as it doesn't bottom out as quickly.

    Regarding the future, I would give direct milling a real go and challenge it. As you have already seen with the stuff I did for your for your prototype, there is real potential to design very specific insoles based on individual needs. I am sure with your clinical hat on you can give Langer a real challenge.

  25. Joe, I'm guessing that the above post is absolutely meaningless to the majority of the people reading this thread- sorry if I'm doing anyone down. So, lets put into real terms that people can understand, what we want to do is predict the thickness of the shell required (y) based on the body weight (x) So according to you: y (shell thickness) = ((4-2)/(200-100)) x body weight + 2-(((4-2)/(200-100)) x 100)

    May ask what data this regression equation is derived from?
  26. Phil, it's probably nothing to do with material thickness, when I've run finite element analyses of models with uniform thickness , the addition of an extrinsic rearfoot post focuses the stresses at this point. End of story. John Weed even talked about this- right, Kevin?
  27. Phil Wells

    Phil Wells Active Member


    Did you ever change the shape of the post - e.g. round off the corners - to see if that changed the stress lines?

  28. joejared

    joejared Active Member

    Hype? It's really only 7th grade math, aka pre-algebra. (y=mx+b) Not much hype to entering a patient weight and selecting a rigidity after the equation is defined.

    What you're describing is what the practitioner should be filling out on an orthotic prescription. (rigid, semirigid, normal, semiflexible, flexible for example) What the laboratory needs to do is deliver the device with the desired flexibility.

    The most likely cause is feed rates and chip-load increase, likely on the trailing edge of the perimeter cutout of the heel. When the system is cutting out an extrinsic posted device, OreTek based systems slow down for extrinsic post's cutout to prevent that problem from occurring. This issue is a basic machining feeds vs speeds concept.
  29. Phil Wells

    Phil Wells Active Member


    If you think that your math is that intuitive that it can extrapolate from patient weight every know factor to ensure you have got the orthotic stiffness exactly required to deliver the correct ORF, then I am putting you up for Nobel prize for medicine (Or literature).
    Seriously your approach is good but it is over kill.
    Just out of interest, how much emphasis do you place on tooling geometry when applying milling strategies? This in itself can be make significant differences due to variables such as (For the laymnan who may be interested) - rake angle, material coatings, speed, thermal changes, coolant type (If any), chip extraction, bull/ball/modified end geometry etc.

  30. Agreed. Moreover, the cost of the man-power involved in a commercial environment with one person reading the prescription, another doing the plaster work, another vac forming, another grinding. I'm with you on that, Phil.
    Agreed, which brings us back to Ed Glaser and his "calibration"

    Now I don't go around hugging trees, but I do find this a little bit frightening. So we are using plastics, not the greenest of products and then we are wasting 75% of it. Now I'm all for "improving my patients footprints, but at what cost to their carbon footprints?"

    Are there any labs using recycled plastic or investigating the use of "greener" materials? Such as bamboo based "plastics"
    Nothing wrong with the stuff you did for me Phil, just the components I put into them that didn't stand up to wear. I'm on the case.
  31. joejared

    joejared Active Member

    y=mx+b (point slope)

    It's only a loose example of how to use a wrk# design code variable. Labs that use this feature define it themselves to suit their own designs by copying it, setting an upper and lower band, and implementing it in the patient's database record for that device. The actual name of the resulting design code would be human readable, such as "RIGID". When rigid is called for, the equation does the work for them, instead of having to refer to a rigidity chart, or worse, guess.

    In OreTek, a design is one or more pages of design code variables with the same human readable design code name. Typically, there are 3 to 4 designs that make up a finished product, not counting those fields entered or transferred from a prescription. There can be up to 17.
  32. Not as yet. I can run it tomorrow and see. My guess is that if you used a high radius fillet or a graded bevel to the distal edge it would reduce stresses here. Personally, I'd remove the straight edge...
  33. Phil Wells

    Phil Wells Active Member


    I admit I am turning a bit 'green' since having a baby pod (now 3 1/2 years old).
    I would love to use more environmentally sound materials but worry about the properties letting the patient down with associated litigation.
    Shame about the prototype - I am trialing something at the moment but the manufacturing cost are prohibitive.

  34. Yeah, thanks. I understand regression equations. I just wanted to know where you got the data from to build the model in the first place.
  35. joejared

    joejared Active Member

    Again, the practitioner would tell the lab what level of rigidity they want. The lab would then enter the rigidity as a code, with some variance based on arch height, such as "rigid", or rigidh", the latter being for a high arch. The implementation is actually never seen by the lab technician. As to whether or not it's overkill, from an operator stand point, it's one less thing to think about.

    My system employs constant climb cutting as the rule, except for wood positives and soft EVA perimeter cutout paths. As the newer systems have spindles that operate at 18,000 RPM, chip load and rake angle become less significant. Older machines with slower spindles do better with cutting angles for aluminum, but that is pretty much an obsolete consideration with the newer and more accurate machines. Coatings prolong tool life, but not enough to matter much, and usually not much more than the extra expense. As for chip extraction, you can see that my newest machine has that option, which works with a standard shop vac or alternatively, a commercial dust collection system such as a powermatic. The dust extraction box can be removed easily by pulling 2 lynch pins and pulling back 1/2" and all flaps have templates to manufacture from standard materials. From the right video on my website, you can see that it performs nicely, and just as good with polypropylene. That's because I designed and built it for the job at hand and the machine/spindle in question, and not off the shelf from Techno. The newer LC3024 based machines from Techno, after I've made my changes, are EVERY bit as good as a commercial machining center 4 times its price.

    I think my photo albums on facebook are publicly available, as are the Youtube videos.
    Last edited: Sep 1, 2010
  36. joejared

    joejared Active Member

    That was purely subjective, from somewhere within the confines of my twisted brain. Those numbers would be closer to flexible than normal rigidity from my perspective, as my own devices for 170 lbs are approximately 4MM thick, but then again, I'm not very normal. ;-)

  37. So, that model is bollocks?
  38. joejared

    joejared Active Member

    Not broken, just not necessarily what a given lab uses. They define the range sometimes with some guidance from me in implementing their ranges. The photo isn't exactly model material, except for maybe the wife beater and soft EVA devices. ;-)
  39. I like the photo Joe. So to be perfectly clear: the regression equation you gave a few posts back to calculate shell thickness was not based on any data and was purely hypothetical? We don't want to mislead anyone now do we Joe? Personally, I would think that body weight would be only one predictor of shell thickness and that we would need a multiple regression model which may or may not be linear.
  40. Thick as **** me, lets look at that equation again: 4-2 = 2/ 200-100 = 1, so 2 /1 = 2.

    2 x body weight + 2 - 2 . So, 2x bodyweight = shell thickness. OH, really? Me thinks not.

Share This Page