Lightweighting Rocker Arms

Here is an example of Lightweighting, as applied to rocker arm design:

http://www.altairenlighten.com/news/renault-uses-additive-manufacturing-to-save-120kg-from-a-4-cylinder-engine/

A big advantage of Additive Manufacturing (aka 3D printing) is elimination of specialized tooling, and minimal post build machining operations. This makes it quick and inexpensive to create prototype and low volume specialized parts (like super light rocker arms for your turbocharged Humber Super Snipe).

Given the proliferation of design tools, and with metal AM printing becoming a mainstream commodity, how long until items like these start becoming available? ( Hello, anyone want to start up a Boutique AM business, aimed at racers? )

If you made an AM rocker arm, and it was found to hit the valve cover, then open the CAD tool, and with a few clicks, the design can be altered to eliminate the issue, the digital file saved, and a new set built right away. No tooling changes, costs exactly the same to manufacture.

Usually called weight savings

As someone who's had an aftermarket LS1 rocker arm snap while driving at 30 mph, strength is a concern of mine. What's the cost per cubic inch of material that's as strong in tension as cast or billet?

Looks like most of the weight savings comes from a skeletonized version of the stock parts, which could be done with cast or AM.

Keith Tanner wrote: Usually called weight savings As someone who's had an aftermarket LS1 rocker arm snap while driving at 30 mph, strength is a concern of mine. What's the cost per cubic inch of material that's as strong in tension as cast or billet? Looks like most of the weight savings comes from a skeletonized version of the stock parts, which could be done with cast or AM.

I believe our 3D prints are similar in strength and fatigue to as-cast, though we print with fairly exotic stuff which may have a weight disadvantage over aluminum from the onset.

In reply to Keith Tanner:

If the new shape you use cuts the peak stress by half, and the material is only 2/3 as strong as a forging, are you ahead or behind?

That's making an assumption about improved strength via design, which may not be possible. You can make the forging stronger that way as well. It was a fairly straightforward question, assuming a magical 50% decrease in stress is a marketing way to avoid answering it

clshore wrote: If you made an AM rocker arm, and it was found to hit the valve cover, then open the CAD tool, and with a few clicks, the design can be altered to eliminate the issue, the digital file saved, and a new set built right away. No tooling changes, costs exactly the same to manufacture.

If this happened, you are doing something wrong. CAD tools should prevent that from happening in the first place. Just like CAD tools should be able to predict the strength of the object.

This isn't a pure trial and error game. The computers should give you a massive head start.

Has there been some magical change in the strength of pot metal in the last decade? I'm not up to date of 3D printing, but I find it very, very hard to believe I can spit a rocker arm out and have it be anything other than a sample to create a real part out of billet or a forging or some such.

I can't imagine anything spit out a speck at a time would equal even a bad casting in strength.

Streetwiseguy wrote: Has there been some magical change in the strength of pot metal in the last decade? I'm not up to date of 3D printing, but I find it very, very hard to believe I can spit a rocker arm out and have it be anything other than a sample to create a real part out of billet or a forging or some such. I can't imagine anything spit out a speck at a time would equal even a bad casting in strength.

We put parts in gas turbines today made that way, the cutaways would blow you away, they are better than most castings, and we typically use pretty neat materials.

In reply to tuna55:

How? Are they done in an inert gas atmosphere? I'm confused about how its done, partly by the oxygen contamination, and partly from the "How the heck do you make a drop of metal stick to another drop that's probably cooler than it needs to be?"

I could build a rocker arm from not much with my mig welder, but I wouldn't count on it surviving.

In reply to Streetwiseguy:

There are some fairly old methods to do that. Good welding does a great job adding material....

The various "3D printing" methods out there are really interesting. I'd love to hear Tuna's experiences, especially at the exotic end. It's not all just pooping little pieces out of a nozzle, the laser sintering tools look like they have real potential but I suspect they're a long way from being a cost savings for most.

why don't you just eliminate the cam and rocker setup and just put the valves on electronic servos? then you've reduced weight and parasitic power loss and can program an infinitely variable lift and duration. Going drag racing? download a power cam, gotta tow? Download a torque cam. Long road trip? Economy cam.

I've been wondering this since I was 12. Surly somebody smarter than me has thought of this already.

In reply to AClockworkGarage:

Koenigsegg did

Linky

AClockworkGarage wrote: why don't you just eliminate the cam and rocker setup and just put the valves on electronic servos? then you've reduced weight and parasitic power loss and can program an infinitely variable lift and duration. Going drag racing? download a power cam, gotta tow? Download a torque cam. Long road trip? Economy cam. I've been wondering this since I was 12. Surly somebody smarter than me has thought of this already.

Big development expense, and a huge electrical power suck, and trying to find a solenoid that will open and close valves 1000 times a minute for years on end, or 4000 times a minute for a short time...

Not insurmountable, but tough and expensive. Camshafts are pretty easy.

Keith Tanner wrote: The various "3D printing" methods out there are really interesting. I'd love to hear Tuna's experiences, especially at the exotic end. It's not all just pooping little pieces out of a nozzle, the laser sintering tools look like they have real potential but I suspect they're a long way from being a cost savings for most.

I cannot comment too much, both because it's not really what I do every day, just some other folks a few hundred yards away, and also because of intellectual property.

Instead, I'll just point you to some publicly available sources which may give you some insight.

http://www.techrepublic.com/article/how-ge-is-using-3d-printing-to-unleash-the-biggest-revolution-in-large-scale-manufacturing/

The end result is an engineering marvel, one monolithic piece that has replicated the complex interior passageways and chambers of the old nozzle down to every twist and turn thanks to the miracle of direct metal laser melting where fine alloy powder is sprayed onto a platform in a printer and then heated by a laser, and repeated 3,000 times until the part is formed. What makes the new nozzle so special isn't just that it has converted a many-steps engineering and manufacturing process into just one. It is also a miracle of material science since it happens to be both 25% lighter in weight, as well as a staggering five times more durable than its older sibling, all of which translates to a savings of around US $3 million per aircraft, per year for any airline flying a plane equipped with GE's next generation LEAP engine, developed by CFM International, a joint venture between GE and France's Snecma (Safran).

http://www.bizjournals.com/boston/news/2016/10/07/viewpoint-from-assembly-line-to-digital-thread-the.html

https://3dprint.com/127906/ge-smart-factory/

http://www.gereports.com/inside-ges-brainy-factory-of-the-future-what-happens-when-you-link-3d-printing-and-the-internet/

This last link is a bit closer to home

Larger than two football fields and emblazoned with a giant GE monogram, the facility is like a massive toolbox from the future. It holds a sleek “microjet” cutter that sends a laser beam through a thin stream of water and cuts shapes into the hardest metals that are so precise they look almost alien (see below). There are rows of industrial-grade 3D printers and ovens with argon atmospheres that cure parts made from light and heat resistant supermaterial called ceramic matrix composites (CMCs). Elsewhere, a robot nicknamed Autonomous Prime after the Transformers character Optimus Prime, scans its work area with LIDAR eyes — the same technology used by Google in self-driving cars — and services a computer-controlled milling machine. Much of the technology here comes embedded with sensors that stream data over secure Industrial Internet links into the cloud for analysis and insights.

...

The type of 3D printer GE Aviation is using is called a direct metal laser melting machine. It uses a laser beam to fuse layers of fine metal powder together and grow the part from the ground up. But Goodwin’s parts were too large and took too long to print on DMLM machines to be economical.

Goodwin told his team to find a more cost-effective machine. Since 3D printing is still so new, the only faster machine they could find was a 3D printer with two laser beams that wasn’t even on the market. “We bought three of those,” Goodwin says.

It was a risky move and problems quickly popped up. “We followed the instructions, but the metallurgy wasn’t good.” His team spent four months analyzing 200 different software parameters and made 400 hardware modifications before they debugged the machines. This spring, they used them to optimize the design of a 3D-printed fuel nozzle for GE’s latest-generation gas turbine and bring it to production. The new design lowers the machine’s nitrogen oxide emissions and increases power output and efficiency. “We were able to run through 10 design iterations in just a few months and then ship the final version into production four months later,” Goodwin says. “Normally it would take us a year.”

alfadriver wrote:

clshore wrote: If you made an AM rocker arm, and it was found to hit the valve cover, then open the CAD tool, and with a few clicks, the design can be altered to eliminate the issue, the digital file saved, and a new set built right away. No tooling changes, costs exactly the same to manufacture.

If this happened, you are doing something wrong. CAD tools should prevent that from happening in the first place. Just like CAD tools should be able to predict the strength of the object. This isn't a pure trial and error game. The computers should give you a massive head start.

CAD tools are no smarter than CAD Users. Unless the user has provided the correct path envelope constraints, the tool has no way to magically guess the right answer.

Of course, the smart way is to build it in plastic first and do some trial fitting, before committing to a metal run.

BTW, it's ALWAYS Trial and Error.

But as you say, never 'purely'. The only issue is by how much.

I've been a hardware and software developer for more than 45 years. And I trust NOTHING 100%. Even testing and verification has failure built in, the best you can do is the best you can do.

AClockworkGarage wrote: why don't you just eliminate the cam and rocker setup and just put the valves on electronic servos? then you've reduced weight and parasitic power loss and can program an infinitely variable lift and duration. Going drag racing? download a power cam, gotta tow? Download a torque cam. Long road trip? Economy cam. I've been wondering this since I was 12. Surly somebody smarter than me has thought of this already.

This is one of the many reasons why manufacturers are looking in to 42V systems. Not only does it reduce parasitic drag and allow infinite valve timing/lift tuning, it can eliminate the need for a starter. The crank sensor tells the ECU what piston is closest to TDC, opens that intake valve, dumps some fuel, and then some spark and you have an instant running engine.

DLMM and SLS are quite nice tech, but extraordinarily pricy; upwards of $500k for small build envelopes. That doesn't include the post-processing and material handling systems either. A couple of us went to CES and peeped this bad boy, which combines fused deposition with metal sintering. It's unproven and something to watch, but costing an order of magnitude less than a comparable laser sintering machine, I can't believe that Markforged doesn't have a line of people waiting to get one. They also make a ~$10k continuous-strand carbon fiber/kevlar/fiberglass printer.

tuna55 wrote: It is also a miracle of material science since it happens to be both 25% lighter in weight, as well as a staggering five times more durable than its older sibling, all of which translates to a savings of around US $3 million per aircraft, per year for any airline flying a plane equipped with GE's next generation LEAP engine, developed by CFM International, a joint venture between GE and France's Snecma (Safran).

WOW.

tuna55 wrote: We followed the instructions, but the metallurgy wasn’t good.” His team spent four months analyzing 200 different software parameters and made 400 hardware modifications before they debugged the machines.

That sounds like my experience with 3D printers. "They're awesome, but you might have to tweak it a bit"

In reply to clshore:

If you don't have the correct location and space laid out, how can anyone make something like a rocker arm with any confidence?? 3D printed parts are not cheap, to the point that I'd spend a lot of time with the CAD hardware to check clearance and strength.

In real engine development, the trial and error part in the mechanical parts is pretty darned small. It's rare that parts don't fit, let alone don't work right away for prototype engines.

Cost in both money and time is too high to not know to a very, very high degree.

The original post seems to imply this is for people who rely more on trial and error than engineering. In the case of plastic 3D printing, you can get away with that as long as you're not in a hurry because the cost of the materials is basically nil. But they have no strength.

So, I read the article and I'm confused. It says that this will "reduce the weight of a 4-cylinder engine by 120 kg or 25%”. So, if 120 kg is 25% of the weight of the engine now, that means a their 4-cylinder engine weighs 480 kg (1056 lb). I think I can cut that by 25% without using a 3-D printer. What am I missing?

It's a diesel truck engine. Apparently used in the D class trucks, like this one.

Keith Tanner wrote:

tuna55 wrote: It is also a miracle of material science since it happens to be both 25% lighter in weight, as well as a staggering five times more durable than its older sibling, all of which translates to a savings of around US $3 million per aircraft, per year for any airline flying a plane equipped with GE's next generation LEAP engine, developed by CFM International, a joint venture between GE and France's Snecma (Safran).

WOW.

Yes. It's literally changing the game. Within a decade manufacturing will be fundamentally different as a result. The biggest benefit was only hinted at in this thread. The low hanging fruit is the headline stuff, the big changes in particular uses of the technology to get big gains because of the ways you need the widget to be shaped.

What the really interesting impact will be is in flexibility. The same machine can now print widgets for very different applications with almost no additional tooling cost. All mass production gets dramatically more efficient, even than it is already. Furthermore things are scale able in ways we never considered. How about the "mod motor"? Imagine the head built only for one cylinder, and you just print what you need, so a I4, V6, V8, V10, whatever, can get the same head design, just scaled. Obviously this is a simplistic explanation, but the opportunities are huge.

Now, as far as cars go, put on your future hat and think about being able to print metal and plastic with the same robot. Now you can make, say, a vehicle speed sensor for a 1988 Honda Accord, something which is NLA outside of a junkyard, for tiny amounts of money if you can come up with a reasonable reverse engineering scenario. That's a niche example, but if a big enough entity, say Summit Racing, wanted to buy a printer and hire a few engineers, they would concierge their way into a neat market.

I think it'll take longer than a decade before very high volume items like cylinder heads will be printed. Your example is a fairly low volume, highly complex wear item that's part of an incredibly expensive assembly, so it's the perfect application.

Digital printing on paper is a good analogy, I think. You can do all sorts of interesting things by personalizing the printout to an individual and get really fast turnarounds along with flexibility of low volumes. But when it comes time to print out 30,000 catalogs, the old school offset press still rules the roost.

Your printed 1988 Honda VSS is a good example of the sort of thing that is well suited to digital production once the tech is there. Now someone just has to do the engineering. That'll be the limiting factor for the same reason that you can no longer buy shocks for a 1985 CRX. The tech is there, but it's not worth putting the investment into the engineering. It might fall to the enthusiast community, but then you get a lot of trial and error as amateurs try to match the durability of the original. By the time the engineering is done, the cost of the actual duplication is not significant.

There's a middle stage as well. You can print molds for casting. Gives you a bit of both worlds - flexibility and complexity but it's less expensive than laser sintering.