• 11

    Giovanni6 days ago 0 comments

    I finally completed the plastic injection mold's BOM and the parts have been ordered. Here it is! If you have any questions, suggestions, I'll be happy to talk about it. This is a screenshot, but all this will be added to the parts list. I have learnt that McMaster is a great source for specific components, but for production runs, their prices quickly add up.

    There is only one thing that has been bothering me a little and it is the mold's alignment when clamping. The Sliding Shoulder Bolts are screwed to the mold's core (top), and they hang on the top platen of the plastic injector. This is what allows the mold's core to slide up and down. Well, there is some tolerance between the shoulder bolts and the platen. It is necessary, otherwise they wouldn't slide smoothly. That tolerance is about 1/64" (0.4 mm).

    The question is, will the tapered interlocks align the mold precisely? I know, there is such thing know ans alignment pins. But I did not add them to this mold. Complicated enough as it is!. And I know, we're talking fractions of a millimeter. But still :) .

    Mental note: Laser cutter are not printers.

    There is another cute little project on the way. Will show more next week!

    This felt like a short, productive week. And the next will be even more intense. So here is this photo from last Thursday to remind you that not everything in Pasadena is work. Cheers!

  • 10

    Giovanni05/14/2017 at 21:53 0 comments

    So I was prepared to start a 3D printing of the whole mold, with cavities other components and stuff. Of course, everything printed separately. I am planning to build a mock-up mold that we can load on the injector and verify dimensions before buying and printing the expensive stuff.

    That's when David suggested a great idea: while there are some components that must be 3D printed, some parts of the mold can be layers of mdf cut with a laser cutter. This probably requires an extra 2 hours of my time, but the mock up core/cavity can be ready in one day, instead of the planned 3 - 4 days of a gigantic printing task (provided that my prayers to our Lady of the Perpetual ABS warping are heard).

    The BOM will be ready tomorrow and I will post it here a few days later. If anyone finds something there that does not look right, any comments will be deeply appreciated!

    As far as the Organizer goes, it is experiencing some manufacturing problems. Apparently, maple's hardness is too much for the band saw and it will not cut through. That forces us to cut a series of steps to form a slope that then can be shaved off. Interesting approach, but if you are ever thinking of mass manufacturing (where you usually pay machine/hour use), this approach is far from desirable. Back to the drawing board!

  • 09

    Giovanni05/07/2017 at 04:20 0 comments

    I love the smell of..... methacrylate resin, burnt mdf, and molten ABS in the morning.

    Smells like... victory?

    All three machines were operational (I was working on the extruder) and it was early in the morning so I wasn't giving up my cup of coffee just to go and make sure my printing was working right. How irresponsible would that be?

    When designing products or molds that need standard components, nothing beats McMaster-Carr and Parts4CAD. It is really easy to look up specific components directly from Fusion 360 and load them to your models. Instant gratification (or frustration)!. The McMaster library definitely works for Solidworks too, and there might be other 3D Apps that can use it too. Or just download the STP files directly from their website.

    So not a lot of photos this week. I have been working on the mold and that means some walking down to the injection machine and browsing through papers and more papers. Sometimes dimensions are expressed in mm, sometimes in inches!

    And this is what's been keeping me busy. Almost all the basics are there, next week for fine-tuning, stock material acquisition, standard components verification:

  • 08

    Giovanni04/29/2017 at 01:46 0 comments

    This week flew fast!

    Shopbot is a few steps away from being fully adjusted and ready for some serious CNC toolpathing (if that's a word). In the meantime, I have moved away from plastic design, to... MOLD DESIGN! Yup, that was fast. Now, at any given time I have around ten thousand tabs open on my Chrome browser, all related to: Advanced Injection Molding, Beginning Injection Molding, to Google, define: "molding" (if you don't get the Simpsons' reference, you're too young).

    The molds plates will be made on aluminum and the cavity and core will be 3D printed. This is new technology that is still being tested, so there are not a lot of documents with definitive answers. But I think the combination of a small plastic injection machine + Tormach + Form 2 SLA 3D printer can only yield interesting results. Not a lot of interesting images this week, but something tells me the next will be pretty exciting.

  • 07

    Giovanni04/22/2017 at 19:32 0 comments

    If you follow the SupplyFrame Designlab's twitter (which you definitely should), you might already know that the ShopBot is undergoing an intense revamp by Dan. That will allow us to have a more flexible work area.

    And not being able to use the ShopBot means... time to use the Formlabs Form2! I am just a few steps away from preparing for plastic injection, and the process I usually follow could be summarized as follows:

    Looks like a lot of steps but unless you know exactly what you are doing, this is what helps me achieve the best results in a relatively short period of time.

    And so we can see a lot of prototyping in the process. FDM is by far the cheapest and quickest option so there's where we study form and function. Lot's of screw ups and fun and games. We make some physical assumptions too. Once the design has been optimized, it is time for the real stuff: To adapt the part for the manufacturing process of choice. In this case it will be plastic injection with 3d-printed cavities. Time to optimize, add draft angles, study plastic flow, etc.

    SLA models are not perfect but they are the next best thing available: Inspect angles, shapes, etc. I thought my model was ready but a few things showed up there. Among them: draft angles made the ribs too thick, so i "lifted" them up a bit, one rib was missing, and the bamboo soap dish that rests on this tray barely fits (I could have noticed this last part with an FDM prototype!). Time to go back to the drawing board and work around the edges. Most Design Software have a draft angle analysis but I have found Fusion 360's to be really easy to use:

    Dave tells me that I should give the 0.01 in tall letters some draft angle but I'm still not convinced. Will I regret not following his advice?

    And the new model is printing now!

  • 06

    Giovanni04/13/2017 at 17:16 2 comments

    Well it turns out that wood is not like plastic. Who would have thought that. If you are buying wood for special projects, it is pretty much like buying apples: You should go to the market, touch them, figure out if this was a good season, and hand-pick your favorites.

    I strongly recommend this 5-minute read with a good, concise explanation about lumber and how it is sold. It will allow you toss a few tech terms at the lumberyard that'll make you look smart. I like to think that's what it did to me.


    Anyways, wood can be bought rough or surfaced. The former is cheaper, but you have to, er, surface it. So make sure you have a surfacer, otherwise it will be difficult to get a smooth finish and most importantly, to properly zero it on a CNC.

    On to packaging design. I had a rough idea of how I wanted the boxes to look, and that's when Maggie came to the rescue! She helped me optimize the box design, and taught me the "slice and dice" technique to properly arrange the components a box must have.

    slicin' n' dicin'

    Yes, I had some Spam Sushi. Not as bad as it sounds, actually. But do not expect to see it in the Michelin Guide anytime soon.

  • 05

    Giovanni04/04/2017 at 18:26 0 comments

    Busy days!

    The first part of the project, having the wood components ready for pre-production is almost done. Today we received an end mill that is one of the few last steps.

    Right after the CNC cutting, A very practical jig made by Dan will help us cut the final part at the correct angle!

    Branding. Now testing different power and speed for the laser cutter/engraver.

    Now on to the plastic components...

    3D printing goes wrong. The Rockstock has been giving us some trouble for a few days. Great printers with a really tall print space.

    .... and the humble Bukito saved the day! It's a bit slower than the Rockstock but it made useful parts that allowed me to figure out a few measures.

  • 04

    Giovanni03/30/2017 at 17:40 0 comments

    Both the Makeup Organizer and the Desktop Organizer are ready now with all the findings after making the samples. Guys, CNC manufacturing optimization is a lot about feeds, speeds, and tool changing.

    Fusion 360 makes life easier as it integrates a CAM module / simulation. This allows to update the design without having to export and move files back and forth. It still can't go to the lumberyard for you though.

    Now it is just a matter of making the last prototype for each one, and on to the short run!

  • 03

    Giovanni03/17/2017 at 01:47 0 comments

    Prototyping week. We product designers fantasize about how easy and friendly we make our products for our manufacturers, specially when we have some degree of understanding of the machines used in the process. Reality hits hard most of the times.

    Your CNC worksheet may be a single sheet of paper, but it is the result of A LOT of work. It details the paths your CNC will follow, feeds and speeds, and tools and bits to use. Here are a few of the tools we used.

    We call this little tool "the beast".

    And here are some of the first results. It is fundamental to adapt paths, fillets and curves to the existing tools available at your shop. How will you know that? Well, simple inventory. Or buy'em as you need'em. Always test toolpaths with scrap material.

    Dan's quick reflexes prevented further damage to the part and the router bit. Holes were bored in a one single step and this can't be good. You need chips and you need to take all that heat out! Next approach: Bore holes in a few steps, or ideally, use standard drill bits.

    This next week we will be refining details on the models based on all the feedback gathered from the first models. And after that, it will be time to have working prototypes ready for the wood products.

  • 02

    Giovanni03/13/2017 at 04:47 0 comments

    So as mentioned previously, this last week we worked a lot on the Shopbot CNC and product development. Dan did most of the CNC legwork. One of the products in the Dejlig line involves organizing women’s makeup and understanding the various sequences involved when wearing it. It turns out I did know way less than I thought. But I got tons of really relevant feedback from my co-residents, and a special shout out to Hunter (https://hackaday.io/hunterfuto) as she really went above and beyond to help me understand quite a few things and gave me interesting tips. Yes I have measured and 3D modeled a ton of makeup, lipsticks, mascara, compacts and what have you.

    One of the main differences between design and development is that development is the assessment and execution from any given idea, to what you can actually achieve with the best possible balance of available resources and expected turnaround. Dimensions, design, processes etc., are carefully studied and optimized so that the transition from idea to reality is smooth. I love this phase.

    The first designs are almost ready, so what is next? You can guess from the next picture: Actual-prototypes-manufacturing! Man I love those words. Can’t wait until Monday!