Wei Shiun Fasteners

What we are

A manufacturer, and also a supplier of Taiwanese-made parts we do not produce ourselves. Both, and it is worth saying plainly rather than picking whichever answer suits the enquiry — buyers ask because it changes who is accountable when something is wrong. For anything in our own range, that is us.

The work sits below where most fastener plants start. M6 and up is a different set of machines: wire, dies, thread rolling and inspection all change as you go smaller. We are not the only factory in Taiwan working this small, and we would rather say so than pretend to a scarcity that a buyer can disprove in one search. What we can offer is a clear answer on what we will take on, at what quantity, and on what terms. Capabilities, including the limits →

The one screw his father made

Ask Wu Chen-wei (吳振維) about his father and there is one sentence: “My dad made one kind of screw for his whole working life.”

That sounds ordinary in this trade and is not. One screw for a working life means the same size, the same dies, the same settings, repeated until you can hear what is wrong without looking. There were no optical inspection machines then. Quality was hands and memory.

The plant was registered before the company — 1986 against 1994 — and the order tells you where this started. Not a company that went looking for work, but machines already running.

Miaoli, Shalu, then north

Wu Jie-wen grew up on a farm in Miaoli. He worked his way through vocational school in Shalu — classes by day and work at night, or the other way round. When school was finished he went north to learn fasteners, and that did not mean an office job. It meant apprenticeship, from the bottom.

He started in wholesale, calling on hardware shops and distributors one at a time, learning who used which size and who was short of what. International trade came later.

That sequence shaped the company. Somebody who came up through wholesale looks at a screw differently from somebody who came up through a factory: the first question is what the customer needs, not what the machines can run.

The tiger is a birth year, not a design brief

The red tiger head has no brand strategy behind it. It is the owner’s zodiac year.

Small Taiwanese firms name themselves like this — a zodiac animal, a place, a child. “Wei Shiun” came the same way: the wei from a son’s name and the shiun from a daughter’s. No marketing department produced it. It is simply what this company is.

We took it because somebody else would not

Asked why customers moved to Wei Shiun, Wu Chen-wei does not talk about price or quality. He gives this example:

Small packs are not a technical problem. Repacking stops a line, takes people, adds an operation — for a plant built on large orders that reads as not worth it, and the answer is no.

None of those four needed better technology. They needed somebody willing to be inconvenienced once. Somebody who came up through wholesale knows what that inconvenience is worth to the customer.

Screws are a jigsaw

Asked how he sees the trade:

That reads like a metaphor and is a description. A screw is not one thing but a set of choices — heads, drives, threads, materials, finishes — and the product of those runs far past anything a catalogue can list.

The square the customer needs is often not on the sheet. Which is why our catalogue is arranged by how you choose rather than by part number.

The screw shops kept my father from going under

In the early years Wei Shiun shipped mainly to screw shops. Wu Chen-wei is blunt about why: wholesale cash flow was steadier and bounced cheques were less likely.

That is not a vision statement. It is small-business arithmetic. Direct customers pay better and pay later, with more risk; the shops pay thin and pay up. His father’s generation took the side that survived.

It belongs here because it explains what came next. Those unglamorous steady years bought the room for the second generation to take a risk.

What the second generation did: making “special” producible

The direction after taking over was development, and the result is this sentence: special small screws, and not one-offs — production volumes.

Instrument makers are a good share of the shipping list, and those parts share a shape: small, tight tolerance, modest volume, and it cannot stop — which is exactly the band an ordinary fastener plant does not want and our equipment happens to fit.

“Can be made” and “can be produced” are two different claims. A machine shop can cut you one special screw. A hundred thousand identical ones, delivered, needs tooling, process and inspection.

That is also why the refusals above could be answered. Taking on trouble only works if you can carry it. The first generation kept the company alive; the second turned “special” into capacity.

“That is as good as it gets”

One customer’s previous supplier could not get black oxide black enough. The answer from their sales was: not possible, nothing we can do, that is as good as it gets.

We sorted it out.

This case differs from the other three. Those were would not. This one was cannot — which is more common and harder to argue with, because the customer has no way to tell a real limit from an unwillingness to try.

Why “not black enough” is a real problem. Black oxide depth depends on bath chemistry, temperature, how clean the part was beforehand and the oil applied after. The coating itself offers very limited corrosion protection — appearance is close to the whole of what it delivers. So a finish that is not black has not achieved the point of the operation.

We could deal with it because we work with a specialist chemical supplier. Most fastener plants outsource finishing, and whether you can move an outside plant is itself a capability.

Note the second half. Not anything is possible, but try before saying it cannot be done. The difference shows up at the first hard problem.

We make it and sell it, and we will tell you how to change the drawing

Wei Shiun manufactures and sells directly. No trading company in between. Two practical differences:

• No margin in the middle. That one is easy.
• The person answering your technical question is the person making the part. Through a trader your question is relayed one way and the answer the other, by somebody who was not on the floor. Drawings are exactly where that hurts.

So we do more than produce to print. We will say this head diameter does not need to be that large, this thread length can be shorter and save material, this tolerance is tighter than your application needs, and this size would be cheaper as a standard part.

That last one is meant. If a standard part solves your problem we will say so. We earn less on that job and you spend less on the wrong thing, and you come back with the next question. A supplier who never recommends a standard part has a “yes” that is not worth much either.

One washer, a million a year

A concrete example: a washer co-developed with a partner, now running a million pieces a year.

• Not build-to-print. Co-developed means we had a hand in what it became.
• Not a sample run. A million a year is a stability problem, not a feasibility one.
• Not only screws. Washers, nuts, rivets and springs are in scope.

We are not writing what it goes into. Anonymising the customer does not anonymise the application — an application identifies a company just as well. Naming it needs their written consent, and the same rule protects you.

“If we don’t have it, I’ll find it”

Asked why customers stay, the answer is neither technical nor about price:

That is the real use of the double role — manufacturer, and supplier of Taiwanese parts we do not make. Not an extra margin. One window to ask through.

If we make it, we make it. If we do not, we find you the plant in Taiwan that does, rather than leaving you to start the search yourself. To a buyer assembling five different parts, that saves more than money.

What you can ask

Fifteen years and more of making screws. These he can answer:

• Whether a size is makeable, and in how many forming stations
• Whether a head change is a tooling repair or a new set
• Whether the plating will stop the thread going together (there is an article on this)
• Whether a standard applies at your size (and on this)
• How large the pilot hole should be, and why the one you have is splitting the boss
• Whether your volume is reasonable in that size band
• Whether changing the screw solves a tolerance stack-up, and what has to be revalidated

And where he cannot, he says so. That is what underwrites the rest — with a supplier who says yes to everything, you cannot tell which yes was real.

Why custom screws happen at all

One of the questions we are asked most: why would anyone need a custom screw when there are so many standard parts?

Early in a design the screw is rarely the problem — pick a standard part and move on. The problem turns up at assembly: the head fouls the housing, the thread bottoms out, the driver will not reach, the countersink angle does not match.

So the demand for custom screws is usually not a wish for something special at the start. It is that late in a design the screw is the last degree of freedom still movable, and affordable to move.

Which is why “what is your minimum order” is the wrong first question. The right one is whether changing the screw solves the problem in front of you, and what that costs against changing the housing. Send the drawing and we can work that out.

We will change your drawing, and say why

“I am somebody who really likes changing customers’ drawings.”

Not fussiness. Fastener drawings are often produced by people who have not stood on a floor — no fault of theirs, but the real limits of heading, rolling, plating and assembly are not visible in CAD. What is perfect on screen does not go together on the line.

• Tolerances tighter than the application needs. Cost and reject rate both rise for nothing
• No allowance left for plating. Especially M1–M1.4, where it will not thread after coating
• A head that fights the forming limit. Head diameter too large against the shank for a single station
• A pilot hole copied from another material. Splitting or no grip, and both happen
• The wrong drive. Whole batches cam out on an automatic line

We point these out and say why. Sometimes the answer is that a change is better. Sometimes it is that a standard part solves it and you do not need us to make anything.

We earn less on the second. But a supplier who never raises a problem is letting you spend money in the wrong place, and you find out at production.

Company

Timeline

Inspection equipment includes vertical and horizontal 2.5D projectors, a Rockwell hardness tester and optical image inspection, alongside automatic packing.

Process

Seven operations, in order:

Steps 5 and 6 are where the size range makes itself felt. Several standards stop above it →

Elected by the trade

Wu Jie-wen, general manager of Wei Shiun Fasteners, served as the ninth chairman of the ROC Screw Development Promotion Association, elected in 2011. The association was then 22 years old with over 150 member companies.

It is here because it is the one credential on this site that did not come from us: named, dated, reported by a business newspaper, and decided by competitors. Commercial Times, 26 December 2011 →

The same report notes he started the business in 1987 doing subcontract screw work, which matches the factory registration of 1986 — and explains the account of a father who made one kind of screw for a working life. Subcontracting is making the same screw, repeatedly, for someone else.

Markets

North America, Europe and the Nordics, Japan, and Taiwan. Each asks for a different set of paperwork, and the paperwork is usually what holds a first order up rather than the parts. What each market asks for →