Engineered Mechanical Systems

Directions Website

Engineered Mechanical Systems Engineered Mechanical Systems started in 1990 in response to the need by other manufacturers to have

Engineered Mechanical Systems started in 1990 in response to the need by other manufacturers to have quality built equipment in a timely manner. As the customer base increased, Engineered Mechanical Systems Inc. diversified into different areas of fabrication and manufacturing to better serve our clients and not become limited to one industry. As we continued to grow, a sales force was added to target areas not yet reached by our company. In May of 2000, we moved our operations to 118 Parmenas Lane. We started with a mere 25,000 square foot facility and since then have expanded to 120,000 square feet with our latest addition in the fall of 2017. This has allowed us to provide a more strategic location for our customers and our employees as well as allowing us to expand our capabilities and add more machines to our shop.

Most aluminum porosity is decided before anybody strikes an arc. That's a frustrating thing to hear if you've spent the ...
10/10/2026

Most aluminum porosity is decided before anybody strikes an arc. That's a frustrating thing to hear if you've spent the morning adjusting the machine.

Aluminum grows an oxide skin. That skin melts far hotter than the metal underneath it.

Strike through a dirty or damp surface and the arc is fighting two melting points at once. The oxide refuses to go anywhere. The parent metal is already going. That fight is where the porosity comes from.

So prep does more for the weld than any setting on the machine.

Acetone, to strip oil and water vapour. A stainless brush kept for aluminum and nothing else, so you're not working steel particles into the joint. And full drying before you strike.

Nobody's ever been impressed by a clean brush. It's still the thing that decides whether that weld comes out porous.

https://engineeredmechanicalsystems.com/welding/tips-you-need-to-know-for-welding-aluminum/

Two tolerances sit side by side on a print and look like the same kind of thing. A tolerance on a diameter, and a tolera...
10/09/2026

Two tolerances sit side by side on a print and look like the same kind of thing. A tolerance on a diameter, and a tolerance between diameters.

They're very different purchases. The difference is what happens to the part in the chuck.

A tight tolerance on a single diameter is bounded work. Tooling, measurement, a finishing pass. The shop knows roughly what it costs before it starts. So can you.

A tight relationship between two features is another animal. Concentricity. Runout. A face square to a bore.

These aren't questions about how accurately something can be cut. They're questions about whether the part can leave the chuck at all. The moment it's re-gripped, that relationship gets re-established from a new datum. Holding it across the change is the expensive event.

It's also specified by reflex, routinely, and understandably so. A concentricity callout takes five seconds to add. It looks responsible on the print. It never looks like the line item it turns into.

So spend the relationship tolerance where features genuinely have to run true to each other. Everywhere else, leave it in the general tolerance block, where it costs what it should.

https://engineeredmechanicalsystems.com/metal-fabrication/turning-design-rules/

Splitting a part in two is usually a machine's decision, not yours.Something somewhere couldn't take it whole. Then that...
10/08/2026

Splitting a part in two is usually a machine's decision, not yours.

Something somewhere couldn't take it whole. Then that decision follows the part forever. It adds a joint, a fixture, an inspection and a tolerance stack that didn't exist before.

So measure against a real envelope before you commit to the split.

This is what this floor takes:
— Plasma: 5-axis, 10-foot by 20-foot tables, up to 4 inches thick in carbon and mild steel
— Press brakes: up to 353 tons and 240 inches of bending length
— Turning: up to 17.72 inches in diameter and 48 inches between centers
— CNC machining tolerances down to ±0.0005 inches, depending on material and part geometry

If your part sits inside all of that, the joint you were about to design in might not need to exist.

https://engineeredmechanicalsystems.com/get-a-quote

Ferrous means it contains iron. That one fact carries the whole trade-off. Iron alloys are dense, strong and cheap. They...
10/07/2026

Ferrous means it contains iron. That one fact carries the whole trade-off. Iron alloys are dense, strong and cheap. They also corrode.

Two definitional numbers live inside that family and they're worth keeping straight. Carbon steel runs up to 2.1% carbon. Stainless earns its name from chromium, at a defining threshold of 10.5%.

The habit worth breaking isn't about the numbers, though.

Most material decisions get made by naming the metal first and justifying it afterwards. Usually it's what the last part was made from. That's a reasonable shortcut, right up until the application changes underneath it.

Name the property first. Strength, weight, corrosion resistance, cost. The metal tends to pick itself.

Dross lands on quotes as an automatic grinding charge, as though it were a tax on the process. It isn't. It's a symptom ...
10/06/2026

Dross lands on quotes as an automatic grinding charge, as though it were a tax on the process. It isn't. It's a symptom of how the cut was run.

There are two kinds and they have opposite causes.

High-speed dross comes from cutting too fast. The arc lags behind the torch. Low-speed dross comes from cutting too slow. The arc shoots out ahead.

Same rough beard on the underside of the plate. Two corrections pointing in opposite directions. Which is why a shop that reads the edge is worth more to you than one that grinds every edge by default.

The heat-affected zone works in a similar way. Everybody worries about it at the cut, which is where it does the least harm.

You feel it one operation later. A bend line running across an affected edge isn't the same bend as one in parent material. Drilling or tapping a hardened edge is harder on tooling and less predictable than the same feature an inch inboard.

The remedy costs nothing at the design stage. Keep bend lines and tapped holes off the as-cut edge wherever the geometry allows.

https://engineeredmechanicalsystems.com/metal-fabrication/plasma-plate-cutting/

Bend radius. Three different numbers wear that one label.Inside, outside, centerline. The print says radius, and now a b...
10/05/2026

Bend radius. Three different numbers wear that one label.

Inside, outside, centerline. The print says radius, and now a bender has to stop and call you.

Specify centerline radius. CLR, measured to the tube's centerline, is the number a bender works from.

Better still, state it as a multiple of outside diameter. That's the D of bend. A 2-inch tube on a 4-inch centerline radius is a 2D bend. Every downstream consequence scales with that one ratio.

General industry guidance from Xometry puts the working floor at 2D. It recommends whole-number multiples. 2D, 3D, 4D. The radius lands on tooling that already exists instead of on a custom tooling charge.

And standardise while you're in there. One radius across a part runs in fewer setups than three radii that each looked proportionate on screen.

https://engineeredmechanicalsystems.com/precision-forming/tube-bend-design/

Change a feature before the drawing is released and it costs nothing. You change it and get on with your day.After relea...
10/04/2026

Change a feature before the drawing is released and it costs nothing. You change it and get on with your day.

After release, the same change is a revision. A new number. A new print. A round of conversations with everyone already working from the old one.

Same edit. Completely different price. The only variable is when the conversation happened.

So bring the drawing to a quote and ask which features are driving the setups. We'd rather look at it while it's still easy to change.

https://engineeredmechanicalsystems.com/get-a-quote

A First Article Inspection looks like paperwork on the way to a production run. The way most schedules are drawn, that's...
10/03/2026

A First Article Inspection looks like paperwork on the way to a production run. The way most schedules are drawn, that's exactly what it looks like.

It's also the last point where committing to that run is still a choice.

Everything downstream quietly assumes the design is genuinely in range. The schedule. The material buy. The second operation. The customer's date.

FAI is where you find out whether that assumption holds. Treat it as a formality and you've spent the money without buying the information.

The paperwork side works the same way round. CMM data is what relieves the bottleneck in a PPAP submission. The dimensional evidence already exists in a form the customer can accept. Nobody has to reconstruct it against a deadline.

Both point the same direction. The value of an inspection sits downstream of the measuring.

https://engineeredmechanicalsystems.com/inspections/what-to-do-with-the-results-of-your-inspection/

Long slender turned parts get expensive at a fairly specific point. It's worth knowing roughly where that point sits.San...
10/02/2026

Long slender turned parts get expensive at a fairly specific point. It's worth knowing roughly where that point sits.

Sandvik Coromant's general turning guidance puts a workpiece length-to-diameter ratio of up to 2:1 as normally acceptable supported on one end only.

Past that the part needs a tailstock or a steady rest. That support is a real line in the cost. It isn't a courtesy the shop throws in.

So hold the unsupported length against the diameter early, while the geometry is still soft.

If a shoulder lets you keep the slender section short instead of thinning the whole length, that's usually the cheaper part. It does the same job.

The rates aren't what changed. What changed is what the duty gets charged on.Everyone's arguing about the rates, which i...
10/01/2026

The rates aren't what changed. What changed is what the duty gets charged on.

Everyone's arguing about the rates, which is understandable. They're the number in the headline. But look at the base first.

As of September 2026, the Section 232 duty on steel, aluminum and copper articles and their derivatives is charged on the full customs value of the imported product, regardless of metal content. That took effect 6 April 2026 under Proclamation 11021.

The duty used to scale with how much metal was in the part. Now everything inside the declared customs value sits in the duty base. The cutting. The forming. The machining. The welding. The finishing. The hardware.

All the work that used to sit outside the calculation entirely.

That's why two parts of the same weight in the same steel move so differently. A plain cut blank barely shifts. There was never much beyond metal in its customs value to begin with.

A welded, machined and coated assembly of the same weight shifts hard. Every hour of that added value is now counted alongside the steel.

So if a re-quote has come back well above what the steel alone would explain, look at the base before you look at the rate.

Walked through in full, with the primary sources: https://engineeredmechanicalsystems.com/innovation/tariff-landed-cost/

Address

118 Parmenas Lane
Chattanooga, TN
37405

Opening Hours

Monday 7:30am - 4pm
Tuesday 7:30am - 4pm
Wednesday 7:30am - 4pm
Thursday 7:30am - 4pm
Friday 7:30am - 4pm

Website

Alerts

Be the first to know and let us send you an email when Engineered Mechanical Systems posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Share

Category