Solidworks has a solid-sweep function that Inventor doesn’t have. But does Inventor actually need it? Read on and decide for yourself.
One of the best parts of my current job working for a reseller, is the variety of cool models we get to build or work on for customers. The customer/supplier relationship is for the most part, like a boring-model filter. What I mean by this, is usually by the time the customer wants us to get involved in a modelling task, it’s either because there’s some sort of funky problem with the file, or the modelling is pretty technical and they’ve run out of ideas of how to tackle it. The latter scenario is always exciting because it’s challenging, and we often have to think outside the box of “normal” modelling techniques to crack it.
For some strange reason, one type of machine (that comes in a huge variety of variations) seems to create more of these “how could I model this!?” questions than any other. That super-important fundamental machine is the humble auger, or screw conveyor.
I’m pretty sure it was our old mate Archimedes who cracked this idea way back before Adam was a cowboy. It probably paled in comparison to his theory about buoyancy which had him doing a nudie run down the main street shouting Eureka, but the Archimedes’ Screw was an incredible discovery nonetheless. You see, he cottoned on to the fact that you could take the principle of an inclined ramp, wrap it around a shaft, and rotate it inside a tube to lift water. Ever since then, countless inventors, engineers, tinkerers and probably moonshiners, have recreated this invention over and over again in varying forms, to move just about any medium you can think of.
Archimedes Screw Animation – Credit Wikipedia (Silberwolf)
A few years ago, I was tasked with redesigning the inner workings of a domestic heater that burned wood pellets made from compressed sawdust. The heart of this was a very short steel auger which lifted the pellets from a hopper and dropped them down a tube leading to the combustion basket. The existing design had a host of problems leading to combustion inefficiencies, and the strict emissions laws in some of our cities meant that to keep selling fires, the company needed to seriously improve the design. Having never really worked with screw conveyors or augers at that point, I seriously underestimated just how critical the fine details of that auger design were. It looked simple, but subtle geometric details made huge differences to the performance of the fire. This project gave me a huge appreciation for this fantastic little gadget, and I’ve been fascinated by auger problems ever since.
The idea of this post is really just to discuss a couple of the common modelling issues, limitations of design software, and the manufacturing challenges associated with producing these designs.
As I said earlier, augers or screw conveyors come in a huge variety of designs and variations for various purposes. The ones we will focus on here are:
- Variable pitch augers – for accelerating or decelerating a medium, increasing or decreasing pressure in a medium
- Segmented (split disc) augers
- Feedscrews for moving cylindrical or elliptical objects down a production line
Variable Pitch Auger:
Variable pitch augers are an interesting challenge from a modelling point of view, because the rate-of-change of pitch in the flighting usually needs to be fairly precise to achieve the required result in the movement of the medium. This next little gem of a technique is not one which I can claim credit for discovering, but I want to pass it on because I find it to be a very elegant solution to the modelling challenges with this type of screw.
Picture this, graph the curve that describes your desired movement/acceleration/pressure/flow of your medium, then wrap it around a cylinder, and sweep the flighting section along it. Simple huh!? Ok….. I’ll provide some pictures in case you don’t get what I mean….
Graph the Desired Motion of the Medium
The vertical lines on the graph are at spacings equal to the circumference of the cylinder (i.e. “pitches”)
Wrap the curve around a cylinder using “Wrap to Face” in Autodesk Inventor 3D Sketch environment
Sweep the flighting geometry down the projected curve
If you want to see the process in more detail, I wrote a more in-depth explanation a while ago on my employer’s company blog.
Segmented or “Split Disc” augers:
This type of auger is more common for bigger diameters and is differentiated from continuous-flight type augers by the fact that it is made up of a serious of pitch length segments that are welded together, and to the center shaft (if there is one.) To illustrate the difference, check out these videos:
Continuous flighting:
Segmented flighting:
So every segmented auger manufacturer will tell you that there is a little black book somewhere that has a whole lot of numbers in it which provides the “secret sauce” for creating the flat patterns for each segment. They generally resemble a disc with one radial slot cut from the center to the perimeter, but the exact geometry required to give the correct formed result is often a closely guarded company secret. If you happen to be in the scenario where the “old-schoolers” in your company have always found the geometry by trial and error and now you want to document it in the form of a 3D model, you may find the modelling to be pretty tricky if you want the part to flat pattern. I was recently shown an interesting approach to tackle this which has a nice workflow with many applications. The author has used an interesting workflow where he effectively creates his own adaptive flat pattern which updates when he makes changes to the formed model, but he doesn’t use the standard Inventor Sheetmetal Flat-pattern tool because it has issues with that type of geometry that has non-uniform stretch.
You may wonder why I mentioned this as being a method for split-disc type auger modelling when the author of the above article refers to “continuous flighting,” but you’ll notice that his flat pattern is actually for one segment only. I guess it’s a terms-of-reference thing and he is calling a series of welded segments “continuous.” Anyway, you get the idea…
Feedscrews:
These are the kind of “augers” you would find in just a food production line that deals with bottles or other cylindrical containers which require precise timing or spacing. They are usually made of a hard plastic like acetal and have a very interesting geometry to them. To manufacture them, a spinning milling cutter moves along the rotating stock cylinder, parallel to the axis of the stock. While easy to visualise, this is not an easy thing to model in Autodesk Inventor. The main limitation here is that Inventor doesn’t have a solid sweep function like Solidworks. Although I have heard mixed reports as to how well the Solidworks implementation of it actually works. All the Inventor attempts I have seen have been based on a sweep along a helical path. The problem is, the sweep feature sweeps a 2D profile along a 3D path, but to get an accurate representation of the shape, we really would need to be able to sweep a volume down the path to use this method.
Ok, the above bit probably sounded like a whole lot of waffle, so I’m going to attempt to explain it with a few images.
First we setup a helical sweep and use a circle (of the same diameter as the bottle we are going to convey) as the swept profile.
Setup of Swept Cut Feature in Autodesk Inventor
Which leads to a shape something like this:
3D Swept Cut with Bottle Shape Superimposed
Now here you can see the problem. The “bottle” actually clashes with the geometry of the screw. This is not the correct shape. Using Inventor’s “Analyze Interference” tool, we get a clearer picture of the problem:
Analyzing the Interference Between Bottle and Screw
If only we could somehow sweep the “bottle” along the helical path as a cut, we’d have our solution, but we can’t, so….. we need another method.
I had all but given up on this problem, and forgotten about it for a few months, when the D&M author you all know and love, Scott Moyse, stumbled across a forum discussion about exactly this challenge. Many people had come up with attempts at solutions, but only one seemed to get the geometry correct, with a fairly elegant solution. Neil Munro is a name that many of you will recognise, and something of an Inventor legend. He came up with a very clever solution to creating the geometry using all sorts of crazy projection. I hate to say it, but even after analyzing his model, I still don’t understand the steps he took to get there. But it seems to work!
Neil Munro’s Solution to the Problem
Neil has even included his model files within his post on the Autodesk forum, as well as a video to show the end result.
So what do you think? Does Inventor need the solid-sweep feature, or is a method like Neil’s suitable as an alternative?
Even if these examples don’t help you in your day job, I hope you find the concepts interesting. It’s all about Design, and Motion.
[subscribe2]