Webster Engine re-design & (hopefully) build:

[lucky13dave]

Hello everyone,

I recently joined this forum, have posted in the welcome section:
http://www.homemodelenginemachinist.com/showthread.php?t=24710

A brief intro here, I love making stuff, have been a machinist (mostly CNC) for about 10 years now. I studied Engineering Management and have a degree. My employer is training 3 of us to help out with engineering tasks. We are working on a Webster Engine as an exercise.

Each of us built the engine in Solidworks, more or less to the original prints. We made dimensioned and toleranced prints, and now they're approaching it like they do a customer. How can we make this functionally the same, but make it easier (and cheaper) to manufacture?

It's been quite a fun way to learn, and they worked in quite a bit of the business aspects that I haven't known before.

That being said, Joe Webster's engine is a great jumping off point. It's a design that is simple and has been proven to work.

 

[kquiggle]

That sounds like a great project. I'm in the process of (slowly) building a Webster myself. In the process of doing so, I collected quite a few links from others who have done the build, and made some improvements (or at least changes) which may be of interest to you (click on the link below and scroll to the bottom to see a list of related links).

Good luck with your project - I hope you will keep us updated on your progress.


https://sites.google.com/site/lagadoacademy/machining---lathes-mills-etc/build---webster-engine

 

[lucky13dave]

Great website, thanks for sharing! I especially enjoyed the "Advice from a Newbie" section.

"But when it comes to newbies, experienced practitioners sometimes have a common failing - they have forgotten what it was like to be a newbie, and they sometimes assume that everybody knows things that have become second nature to them."

Spot on. Sometimes I catch myself with that failing as the experienced guy, and other times I'm on the other end.

Looking forward to your completed engine!

 

[lucky13dave]

From JB Websters Site:


Some of the main design considerations were:

(1) General ease of construction with minimal tooling (basic lathe and mill work).
(2) No fancy radiuses, spokes or bolt patterns requiring the use of a rotary table.
(3) "Smallish" proportions to limit material costs.
(4) Limited number of parts for speed of construction.
(5) Availability of gears, bearings and other components.

Our design considerations are similar. However, we take into consideration our capacities, capacities of outside processes we have available (and usually a relationship with) e.g. plating, heat treat.

As such, our approach to items 1 and 2 are similar, but take into account our capabilities. We're not going to order a new machine, we *MAY* make special workholding, we will try to design around special cutters (the model taper tap for the piston oiler) but will do a cost/benefit analysis before ordering special tooling. Our CNC capability does give us much greater flexibility on features such ad complex profiles and bolt patterns. I expect that our final design will heavily incorporate such features.

Item 3, we're working off his proportions, and they are all within the working envelopes of our machines.

Items 4 & 5 fall within our engineering processes. Basically, what does it cost to buy vs. make. We could spend a lot of time and money learning to hob gears, and may actually end up making a usable gear. But we can also buy said gear. Which will be more cost effective? If we make 5, probably buying. If we were to scale this up, would it be worthwhile to bring it in house? Sub it out to a shop that specializes in gear hobbing? Order it from McMaster?

We will also be addressing scalability. We will NOT be making thousands, or even hundreds of our final design. I'm hoping we make enough for each of us to get one, and will be ecstatic if we make one physical model. But this is an exercise in our engineering process and our services. So we will explore quoting; single piece pricing, small order, large order.

So, the bottom line? Our engine is likely going to be significantly different from the original. We will use processes that violate much of the original design intent so that we can utilize said processes. I'm hoping everyone will enjoy the build.

Our goal is not the engine itself (but we all want to make one, including the owner and the engineer that is running the training,) but rather the process. I believe that I, and the other members of the team, share that with many of the folks on this forum.

Thanks for reading.

 

[lucky13dave]

Some info on the capabilities of the shop:

Milling:
17 Fanuc Robodrills, various, 10, 14, and 21 tool capacities. 4 with pallet changers. 1 with live 4th axis. years from 1996 to 2013. 30 taper spindles.
2 Mori Seikis, 20 tool capacity, 40 taper spindles.
1 Bridgeport
1 Hurco VM-10, 20 tool capacity, 40 taper spindle, conversational control, & my baby
4 HAAS HA5C indexers, can run on any mill.

Turning:
6 CNC lathes (Doosan & others)
3 B&S Screw Machines
2 or 3 Handscrews

Grinding:
2 surface grinders,
3 centerless grinders
1 Older guy with a great attitude who know the heck outta the older equipment!

CAD/CAM:
Solidworks 2015
Esprit I-don't-know-the-year.

 

[kquiggle]

This is not a slam on the Webster design which I think is ingenious (I'm building one, after all). but there is always room for improvement, so I'll be very interested in what you come up with.

One area in particular which could use some attention is the gears: If you buy them, they are fairly expensive, and if you make them that's a bit of a hurdle for a beginning machinist. The ideal solution would be to eliminate gears altogether, but I don't know if this is possible. My solution (without making major design changes) is to make plastic gears.

Another somewhat tricky operation is pinning the flywheel to the crankshaft.

Well, I suspect I am preaching to the choir.

P.S.

Wish I had access to shop like yours!

 

[lucky13dave]

Kquiggle,

It is a great design, and any design has room for improvement.

As for the gears, I believe we're going to stick with the metal gears. I explored using pulleys, but it was decided that that would require additional parts (tensioner) and probably not be advantageous.

As to the flywheel pinning, We want to eliminate that for ease of assembly. Probably going to use a keyway & key.

As part of this exercise, we are learning and (trying) to use DFMA- Design for Manufacture & Assembly. So we have to take into consideration, if this were to be put into production, who is going to assemble it, and design for the least common denominator. We consider things like Poka-yoke https://en.wikipedia.org/wiki/Poka-yoke the art of idiot-proofing, minimizing steps and processing (e.g. machining steps during assembly- we'd rather design parts ready to assemble)

 

[mu38&Bg#]

Some info on the capabilities of the shop:

Milling:
17 Fanuc Robodrills, various, 10, 14, and 21 tool capacities. 4 with pallet changers. 1 with live 4th axis. years from 1996 to 2013. 30 taper spindles.
2 Mori Seikis, 20 tool capacity, 40 taper spindles.
1 Bridgeport
1 Hurco VM-10, 20 tool capacity, 40 taper spindle, conversational control, & my baby
4 HAAS HA5C indexers, can run on any mill.

Turning:
6 CNC lathes (Doosan & others)
3 B&S Screw Machines
2 or 3 Handscrews

Grinding:
2 surface grinders,
3 centerless grinders
1 Older guy with a great attitude who know the heck outta the older equipment!

CAD/CAM:
Solidworks 2015
Esprit I-don't-know-the-year.

How many hundred or thousand are you making?:hDe:

Greg

 

[lucky13dave]

Greg,

As part of the exercise I expect we will quote it at quantities similar to what our actual customers order, probably a quote for 5, 100, 1000. Maybe more, as as the higher quantities will introduce different production methods. For example, I have a part that is begging to be investment cast (I think) but for our purposes that would only come into play at the higher quantities.

As to how many we will ACTUALLY make? I'd like to see at least 1, and hope we make 4 or 5 (one for each of us, and maybe one for the boss.)

-Dave

 

[lucky13dave]

It's been about 5 months and we've made serious progress. And we have yet to cut 1 chip for this project. We did a lot of cad work, analysis, and such. Long story short, If you were to bring the original Webster design into our shop and ask us to make you one assembly, it'd cost about $17,500. If you wanted 5, they'd be $4000 each. You guys do some impressive work. We modeled all the parts in Solidworks, created 3d models for each part, and 2d prints. One major deficiency in the Webster plans, from our perspective as a contract manufacturing and design company, is the lack of tolerancing on the prints. We created part prints that are dimensioned and tolarenced.

Here's some images from the CAD:

 

[lucky13dave]

And here's our redesign. 1 piece cost is about $9500, and at 5 pieces it's $2,150 per.

Here's some more CAD:

 

[cwelkie]

Fascinating insight into the real world of designing-for-manufacture. Thanks for sharing your efforts. I imagine the process would be a good education tool.
Shows us how we do this stuff for love not money!
Charlie

 

[Brian Rupnow]

Lucky13Dave---I have worked in custom machine design and manufacturing since 1965. Many people don't realize it, but most of the small engines and "things" I build would be beyond the financial reach of most of us if we had to pay a production shop to make them.---In fact, some of the more complex things could almost be traded straight up for a small island in the Caribbean!!! Just think--It stretches us all financially to buy the lathes and mills we use to build the model engines. If we had to actually pay "real world value" for the model engines, we would have to find a cheaper hobby!!!---Brian

 

[kuhncw]

Dave, thanks for sharing this. Interesting redesign.

Did you compare inertia between the original flywheel and the redesigned wheel?

Chuck

 

[Mechanicboy]

Not everyone can have afford to large machines that take all the material out of a piece of aluminum block.
Think: Weight price of a block of aluminum compared with a 1/2 "aluminum plate. Throwing away a lot of material worthy of the price. Also the re-designed WEW engine cost more than a original WEW engine.
It is easier to obtain 1/2" plates than a block of aluminum from scrap dealer.
It's best we keep it simple construction to build motor with minimal tooling equipment and minimal material consumption and time consumption and electricity costs of machine tools. We all folks in HMEM group not allways has a fat wallet.

The pleasure is greatest when the simple created engine works and goes.

 

[Jasonb]

Its interesting to see how it can be approached commercially.

The solid block vs built up is a good example. As said a few bits of flat plate are cheaper to buy. But if paying a realistic working wage the short run time for a CNC machine will have the whole base done while the hobbiest still has his hacksaw in hand just cutting up the materials let alone all the time to machine each individual part. A commercial shop will also be able to resell the vast amounts of swarf they produce which goes some way to recovering material costs.

One only needs to look at the ready made engines from the likes of Bohm to see that mass produced by CNC you can buy a whole engine for probably what it would cost most of us for the raw materials alone

Its unfortunate that savings made for economic manufacture are at a cost to the asthetics of the design.

 

[lucky13dave]

Dave, thanks for sharing this. Interesting redesign.

Did you compare inertia between the original flywheel and the redesigned wheel?

Chuck

Yes, we did. Solidworks allows you to evaluate the mass properties of parts & assemblies, including the moments of inertia, rather easily. The redesign isn't exactly the same, but as I recall it's within 10% of the original.

 

[lucky13dave]

Not everyone can have afford to large machines that take all the material out of a piece of aluminum block.
Think: Weight price of a block of aluminum compared with a 1/2 "aluminum plate. Throwing away a lot of material worthy of the price.

From our position as a commercial shop, the material costs are pretty low. Material cost for parts we would machine account for less than 10% of the sell at price. Hardware (bits we buy from screws to sparkplugs) accounts for a little more, maybe 10-15%. At these quantities, the greatest cost by far is the setup costs, accounting for more than 50%. The reason we could, for the original design, sell 1 at $17,500 and 5 at $4,000 each, or $20,000 is that the setup time amortizes across the run.

With that in mind, by incorporating the cylinder head frame and the 2 support plates into the baseplate, we reduced the part count, the number of setups, & the setup time. The increase in the material cost is greatly outweighed by the reduction in these other costs.

Also the re-designed WEW engine cost more than a original WEW engine.

Design of record (original design):
$17,500 @ 1 unit / $4,000 per @ 5 units
Redesign:
$ 9,000 @ 1 unit / $2,150 per @ 5 units

The pleasure is greatest when the simple created engine works and goes.

Even with the fancy machines and software, we are still eagerly awaiting that moment

 

[Longboy]

Maybe a slight contrarian view..... I would hope that in classes taught that the importance of liaisons between engineering, manufacturing with counterparts in marketing and accounting can reach a consensus in developing a wholesale/ retail goal. I understand that your group handles its area of responsibilities towards the goal and the total success of such is not exclusive to any one department.
It is interesting to hear lucky', how material concepts are developed into products and I'm sure that fellow hobbyists here are proud that your employer or group has picked what is (almost) always a hand crafted item, a model gas engine to explain the process! Thm:

 

[kuhncw]

Yes, we did. Solidworks allows you to evaluate the mass properties of parts & assemblies, including the moments of inertia, rather easily. The redesign isn't exactly the same, but as I recall it's within 10% of the original.

Dave,

I asked because your flywheel axial thickness looks thin out at the OD. What was the logic for stepping the flywheel diameters and thinning the section as diameter increased?

Flywheel inertia helps a small single cylinder engine like this one run more smoothly and idle at a lower rpm. More flywheel inertia is better, within reason of course, for model engines.

Chuck