Posted August 28th, 2014 by Bailey Jones
I started teaching a new Plastics Design course this Fall at ACC. We are using Creo 2 software (formerly called Pro/Engineer) as we study principles of injection molding. We’ll get to look at real-life example parts and also will take a field trip or two to local injection molding facilities. Did you know that Tasus Corporation in Georgetown manufactures large, high-end parts for the Toyota Tundra factory in San Antonio? We’ve toured the Tasus facility in previous semesters.
Here’s an excerpt from the Syllabus:
This course builds upon the concepts presented in DFTG 1433 (Inventor) and DFTG 1429 (Solidworks). The Solidworks class is not a prerequisite, although it is helpful to take it before or concurrently with this class. This class covers advanced CAD methods with an emphasis on cast and injection-molded part design.
For this class you should have prior 3D modeling experience and be comfortable drafting to ASME Y14.5 2009 standards. We will be using Creo 2 software to take your CAD expertise to the next level. We will quickly review topics from your previous classes and apply them to this software. Then, we will move past working with basic parts, assemblies and drawings, and begin learning more advanced techniques such as:
- top-down design (skeleton modeling)
- complex surfacing
- cast and injection-molded part design (plastics design)
- design standards for mass-production
- prototyping methods, including the 3D printer we have in our department: a Dimension FDM (fused deposition modeling) machine.
Posted July 29th, 2014 by Bailey Jones
There’s a difference between designing for a prototype and designing for high volume manufacturing. Often it will make sense to start out with a simple CAD model to prove out a design idea. This CAD model can then be 3D printed to produce a prototype. It might be a functional prototype (to show it works) or a visual model (to show that it looks right).
It is important to know that a plastic part that can be 3D printed may be very different than a plastic part that can be injection molded. To move past the functional or visual prototype we must design for manufacturability (DFM). For an injection molded part we must consider:
- nominal wall thickness
- ribs and rib thickness
- minimum feature size
A CAD model that accommodates these requirements quickly becomes more complicated than our original CAD exploration. The hard work it takes to create a great CAD model will pay off with cheaper, more beautiful parts and less expensive tooling.
Have a look at this video that shows a functional prototype CAD model, and then the more comprehensively designed CAD model for injection molding.
Posted November 22nd, 2013 by Bailey Jones
I often get inquiries from individuals who want help designing a product, when it would behoove them to first design a business. The point being that even a great product has abysmal chances for success if there is not a good sales and marketing engine behind it. Design and manufacturing has gotten easier these days as it has become more accessible to more people. And it can be tempting to overlook the even-more-important aspect of business development. So, I find myself sending out emails like this:
Sounds like your first step would be to develop a business plan; that is, determine who your customers would be, determine if the market size for this product is sufficient, develop a marketing strategy, and understand your sales and distribution channels. The biggest point (and it can be tricky) is to validate your market before spending much money.
What we specialize in usually comes next. We take the idea and design a product that can be economically mass-produced. Let me know if/when you are to this point. Hope this helps!”
There are different strategies for market validation depending on the product type. In the software or web-service field it could be as simple as constructing a landing page and asking for email addresses, or releasing a limited beta version. Physical products can be a bit more difficult and sometimes a description in words isn’t enough. The truth is, you may find that you do need a bit of product design early on to produce a pretty picture (product rendering) or even an appearance model or prototype to convey your vision.
But, at this stage of business development, the design of the product should serve the development of the business – not the other way around. After all, you may find that the market is invalidated, in which case you should abandon the project. I don’t have the secrets for market validation and it can be a tricky endeavor, but it is crucial for reducing your monetary risk.
This article was originally written for Pyragraph.com
Posted July 30th, 2013 by Bailey Jones
This article was originally written for Pyragraph.com
I recently designed some small pieces of furniture for a company called Staandup Desk. These stands are meant to go on top of a regular desk to allow you to work standing up. I had a conversation by email with the owner, Amber Hagopian, about her manufacturing experience here and abroad. I began with a question about communication.
Bailey: I’ve found that increasing the opportunities for effective communication usually increases the chances for successful product fabrication and delivery. That’s one reason fabricating close to home, versus overseas, can be an effective option. Do you have any comments about this that you can relate specifically to Staandup desk products?
Amber: I would definitely agree with this. Initially it seemed like we could just send over a purchase order with some drawings to a factory but what we found after sourcing in China and then back in the USA was that there was very valuable input on design and materials from the USA based fabricators that we did not get from the factories in China. This was due to several reasons, one being that the USA based manufactures were more willing to work with a new company and provide the input and secondly there was still a language barrier even with English speaking coordinators overseas. Our products have gone through several design tweaks based on suggestions and recommendations from our states’ based manufacturers which has made our products better quality, lighter in weight and more cost effective.
Bailey: What were your biggest hurdles related to manufacturing your first product in China?
Amber: The biggest hurdles we faced fabricating a new product overseas was reaching the minimum order quantity and feeling confident the factory we chose was going to provide what we ordered and get it to us on time and in good condition. We did not use a USA based third party company to choose a factory and ended up with a 20 foot container of product that we ended up paying to have disposed of because there was no quality control in place. The materials used for actual production were much cheaper quality than the prototype the factory originally provided and we were unable to sell the items. We did not have proper agreements or warranties in place and ended up taking a total loss with no recourse against the Shen Zen based company. In addition to taking the loss we ended up without inventory to supply the demand we created for our product which has been an ongoing issue. For one product we cannot get fabricated cost effectively in the USA we are now using a states’ based company that sources factories overseas, provides quality control and coordinates shipping to our distribution center.
Bailey: What were the biggest advantages?
Amber: The only advantage we have to sourcing overseas is cost. Even with shipping cost, materials and labor are so much cheaper over there that if we hadn’t initially sourced overseas we would not have been able to bring the product to market because there was not enough margin to make it worthwhile.
Bailey: How would you compare/contrast that to your manufacturing experience with the current product in the US?
Amber: Here in the states we have been able to order smaller quantities of product as we tweak our design. This also helps with cash flow as we do not have to place large minimum order quantities and then pay to store them as we ramp up our business. Additionally, the USA based manufactures have been very helpful in perfecting the product design for aesthetic and cost. Lead time for smaller quantities is much less in the USA and being able to be in communication over the telephone with the fabricators here has been helpful.
Posted March 22nd, 2013 by Bailey Jones
I just gave a talk about rapid prototyping at the Austin Hardware Startup Meetup. We hear a lot about 3D printing these days. Rapidly diving costs in this industry are fundamental to all the attention it is getting. Make Magazine recently came out with an edition that lists about 15 different 3D printers (all are FDM machines) at around $2000 apiece. I’ve been using rapid prototyping since the 90s when the machinery cost closer to the top end of 100′s of thousands of dollars. We now have many different rapid prototyping choices at a more approachable cost. I’ve watched drastic change in this industry, and even helped design a few 3D printing machines over the years.
We have a dizzying array of acronyms to choose from when it comes to picking a rapid prototyping method (and these are just a few of the most popular):
- SLA, Stereolithography – laser cured light sensitive resin
- SLS, Selective Laser Sintering – laser sintered nylon powder
- FDM, Fused Deposition Modeling – hot extruded plastic
- RTV, Room Temp. Vucanization – Cast Urethane in silicone molds
- Polyjet – uv cured light sensitive resin, placed with printheads
- CNC Machining, subtractive process by computer controlled milling
If you are tinkering around by yourself, FDM is the way to go. If you have a membership to Techshop, you’ll have access to a Makerbot FDM machine and computers with CAD software (Autodesk Inventor). This machine prints by extruding a bead of plastic though a hot nozzle.
Ideally, the prototyping method would be chosen according to the objectives of the prototype. There’s a wide range of materials from accurate and fragile to durable and less accurate. The materials available depend on the production method. Here’s a chart mapping out some of the characteristics of these methods.
The other side to 3D printing is to generate the 3D computer file that the machines print from. This can be a complicated task. One option is replication, that is to scan and digitize existing objects. More interesting to me is the creation of new things. This requires CAD software. There are a few inexpensive or free options available as I show in the next chart. Have a look at the chart as a starting point for orienting yourself in this CAD landscape. Also, find a PDF of these charts here: Rapid Prototyping Primer
Posted February 26th, 2013 by Bailey Jones
Old and new companies alike are bringing manufacturing back to the United States. Whirlpool now makes a kitchen mixer in Ohio that they previously made in China. Wham-O has done the same with a Frisbee that they now make in California. And GE has two new assembly lines and a new plastics manufacturing plant in the previously abandoned Appliance Park in Louisville, Kentucky with two more lines in the plans for this year. These strategic and financially motivated moves indicate a shift in the offshore manufacturing trend that began in earnest in the ‘70s and became a forgone conclusion by the ‘90s. Why is it that it now makes sense for these companies to manufacture in the United States?
GE CEO Jeffrey R. Immelt explains in a Harvard Business Review Interview in March 2012:
“Shipping and materials costs were rising; wages were increasing in China and elsewhere; and we didn’t have control of the supply chain. The currencies of emerging markets added complexity. Finally, core competency was an issue. Engineering and manufacturing are hands-on and iterative, and our most innovative appliance-design work is done in the United States. At a time when speed to market is everything, separating design and development from manufacturing didn’t make sense.”
That is, having manufacturing, marketing, design, and engineering in the same place increases the chances for the success of an innovation product.
A December 2012 article in The Atlantic Magazine gives an excellent overview of GE’s recent manufacturing shifts and the climate and rational behind those shifts.
The impetus begins with global economic trends as stated in the Atlantic:
- “Oil prices are three times what they were in 2000, making cargo-ship fuel much more expensive now than it was then.
- The natural-gas boom in the U.S. has dramatically lowered the cost for running something as energy-intensive as a factory here at home. (Natural gas now costs four times as much in Asia as it does in the U.S.)
- In dollars, wages in China are some five times what they were in 2000—and they are expected to keep rising 18 percent a year.
- American unions are changing their priorities. Appliance Park’s union was so fractious in the ’70s and ’80s that the place was known as “Strike City.” That same union agreed to a two-tier wage scale in 2005—and today, 70 percent of the jobs there are on the lower tier, which starts at just over $13.50 an hour, almost $8 less than what the starting wage used to be.
- U.S. labor productivity has continued its long march upward, meaning that labor costs have become a smaller and smaller proportion of the total cost of finished goods. You simply can’t save much money chasing wages anymore.”
So, they decided to try make the GeoSpring water heater (formerly made in China) in Kentucky. It turns out that the Chinese model was a manufacturing mess. Kevin Nolan, Vice President of Technology, says, “We really had zero communications into the assembly line there.” So they got together a team including factory workers in Kentucky and redesigned it.
They eliminated parts and their material cost dropped 25 percent. The time to manufacture it dropped from 10 to 2 hours. The quality and energy efficiency improved. The overall time to market improved by many weeks. And, the price dropped from $1599 to $1299.
A misleading allure to overseas manufacturing is the quote for services. That overseas quote will likely be several times cheaper than a competing quote in North America. However, many times I have seen how that margin can disappear by the time the product is received. Time delays, scrapped parts, travel, and miscommunication can contribute to enormous headaches and financial pain during the manufacturing process. In 2010 Harry Moser, started the Reshoring Initiative to evaluate what he calls the “total cost of ownership”. He contends that the savings from manufacturing overseas has been vastly overestimated for many years. The Total Cost of Ownership Calculator on the organization’s website attempts to put a dollar amount to these often overlooked costs.
An Inc, March 2012 article explains an Accenture study that supports this analysis where “the researchers noted a significant underestimation of overseas manufacturing costs.”
“Our study found … that many manufacturers who had offshored their operations likely did so without a complete understanding of the ‘total costs,’ and thus, the total cost of offshoring was considerably higher than initially thought,” concluded John Ferreira and Mike Heilala, authors of the report. “Part of the issue is that not all costs of offshoring roll up directly to manufacturing; rather, they impact many areas of the enterprise.”
“This overreliance on direct costs to the exclusion of other legitimate cost factors distorts the business case for offshoring, and likely many decisions to offshore were incorrectly made.”
Smaller companies are also recalculating their manufacturing locations. On January 9th , 2013 Marketplace Tech reported on the robotic toy company Cubelets. Founder Eric Schweikert tells of a recent trip to China, “While I was on that visit, we interviewed, I think, six contract manufacturers who would make the entire product for us. Because common wisdom says that when you going into high volume consumer electronics, you have somebody in China make your stuff for you. It was a great visit. We met all those people, and we came back and we decided we’re going to do this ourselves because it’s absolutely insane when you really stop to think about it to make toys all the way across the world. So right now our team is conducting a deep analysis about whether we can build a giant factory outside of Boulder, Co., and hire American labor and build everything here ourselves.”
For 3D Robotics, local manufacturing means manufacturing across the border from their San Diego headquarters in Tijuana. Company partner Chris Anderson, former editor of Wired, explains the idea of “quicksourcing” in a January 26, 2012 article in the New York Times. When they started the company three years ago they produced everything in China, but now they have a second factory in Tijuana. Anderson explains several reasons for this shift in their company.
1. The short supply chain allows them to manufacture in smaller batches. Their orders from China require large volumes of merchandise that remain static as they are slowly sold throughout the year. Smaller production cycles allow more frequent product innovation, and smaller upfront purchases.
2. There is less risk. They can more easily fix any problems of the design without risking large faulty production runs, and they have more control of inventory. Also, there are fewer potential leaks for intellectual property.
3. It is faster. Communication is delayed and even urgent shipments take a lower priority for these factories that serve large multinational clients. He says that they have consistently underestimated the time it takes to get merchandise.
4. At almost $6 an hour, wages in Southern Chinese cities are now almost as high as those in Mexico. Chinese wages have more than tripled in the last decade.
The New York Times continued its reporting on Mexican manufacturing in its February 24, 2013 article “How Mexico Got Back in the Game.” Mexico has more free trade agreements than any other country in the world, graduates a large quantity of engineers and has a vibrant tech start-up community especially around the manufacturing hub of Monterrey. They report that Mexico is also taking back manufacturing market share from Asia. Mexican manufacturing has become more productive than their Asian counterparts and over the past decade Mexican businesses have become increasingly globally minded.
We have lost some of our manufacturing infrastructure in the US. Everything you need to build a bicycle you can find within a small radius on the island of Taiwan for a good price. That’s not the case in the US anymore. Need a specialty screw for your electronic device in China? There’s probably a plant nearby that will make it for you. But, there is a division of Foxxcon in Juarez making Dell Computers. There’s a plant in Louisville, Kentucky making home appliances. And there’s countless Kickstarter projects being sourced and manufactured in the USA. Take a look at Austin’s own SuperMechanical and their product Twine for example. It takes a careful study to determine where to manufacture. We can no longer assume that overseas manufacturing is best for our businesses.
Posted September 21st, 2012 by Bailey Jones
In this post I’ll go behind the scenes for the design of a handheld diagnostic tool. The enclosure had to contain the circuit board, provide a graphical display and have two input buttons. After finalizing the conceptual design, I created the detailed design with Pro/Engineer. This software creates the files that can be used for prototyping and manufacturing. This video gives an overview of the CAD processes used to create the parts. I discuss implementing draft, which is necessary for the injection molding process, and creating a realistic image rendering.
An important part of plastic design is to incorporate draft into the parts. Draft is the slight angle on the vertical features that allows the part to slip out of the tool. The image below shows a draft analysis of the top part. The two different colors represent the surfaces of the two different sides of the mold.
Once all the individual parts have been modeled, they are all put together in an assembly. In the assembly you can check the fit of the pieces, test for interferences and evaluate the device as a unit.
Posted August 22nd, 2012 by Bailey Jones
Here I’ll go over the design process to create a custom cardboard box. This project was to create a 4-gallon container for liquid beverage concentrate. It should be easy to carry and easy to invert and insert into the dispensing equipment. This video steps through the process.
In production, this box will be cut with a cylindrical die cutter. The normal maximum capacity (or maximum box flat pattern size) for the die cutters is 66″x 110″. The arrow below points to the die.
For a prototype, however, the box will be cut on an automated x-y table with a reciprocating blade as shown in the video. The machinery also scores the cardboard at the fold lines. All the data for the contour and fold lines is in the 2D drawing. This electronic file can be input directly in to the machine.
Posted July 6th, 2012 by Bailey Jones
This is the latest prototype (with the modified geometry) on the train. It is made by Selective Laser Sintering (SLS). The material is a fire retardant nylon approved by Federal Aviation Regulations, FAR 25.853. The video shows how it is snapped on and how it holds the bicycle in place. In its final form the fitting would be meant to stay on the rack at all times (although it it easily removed.)
And here’s some pictures of the fitting in place.
Posted May 22nd, 2012 by Bailey Jones
I’ll give a quick review from last week. I tested the prototype fitting on the train and found that the angle of the wedge was off and that the wings of the wedge made it more difficult to remove the bike from the rack. The snapping action that affixes the fitting to the existing bike rack worked very well.
I’ve now modified the original prototype to address the first two issues and I have tested it on the train. The snapping still works very well, the wedge is angled properly so that it fits against the wheel, and the shortened wings make bike removal easier. You’ll see in the photo how I’ve sawed the wedge off and screwed it on in the new position to make a new, Frankenstein prototype.
The fitting does prevent the bike from falling over and it reduces the swinging (but does not eliminate it.) The cord is there so that I could remove the fitting without yanking off the screwed-on wedge. Here’s the fitting in the train:
Finally, I have updated the CAD and produced a rendered image as a preview to the actual part.
Here are some pictures showing how I modified the original prototype: