At Andy Marine, we believe that practical manufacturing knowledge is one of the most valuable assets built over time.
Inside our workshop, there are engineers, production specialists, and experienced technicians who have spent years working with stainless steel marine hardware — from material selection and casting processes to finishing and quality control.
Many of these insights are learned through daily production experience: understanding why a component fails, how different materials perform in marine environments, and what details truly matter when developing reliable hardware.
However, this practical knowledge is often shared only within the workshop and rarely reaches customers who are making sourcing and engineering decisions.
ENGINEER TALKS is our effort to organize these conversations and share the manufacturing experience behind marine hardware.
Through this series, we discuss materials, production processes, design considerations, and real-world lessons from the perspective of a marine hardware manufacturer.
Today's ENGINEER TALKS Ep.40 is presented by Engineer Zhang, written by me the editor.
Two Bases, One Drawing, Two Outcomes
Suppose two cast bases from one order reach the same bench on the same drawing and paperwork. On the first, the stud bottoms out in the blind hole before the joint closes and the gap stays open. On the second, the thread strips and the stud turns free before it clamps anything. Both passed goods-in.
The instinct is to blame one bad casting and one bad fitter. The drawing does not support that split: it gave both the same instruction and the same silence.
So the question is not who is at fault, but what an engaged thread carries, and what the drawing would have had to say for either failure to be caught at the desk.
A Thread Is a Load Path With a Short Memory
An engaged thread looks like identical turns sharing a job evenly. It does not: the first turn does most of the work, and the turns behind it add progressively less — capacity rises with length, but at a falling rate.
"Deeper is safer" also stops being true at the top end. Push it far enough and small pitch differences use up the clearance, so the joint binds before it tightens — which reads on the bench as a damaged thread. Galling belongs to the episode on stainless steel thread galling; this is only why engagement has a ceiling.
Which Member Gives Way
"Threads" is not one part. A bolted joint can give way in three places, and only one is the bolt breaking: the fastener can fracture, the bolt thread can strip, or the thread in the part can strip. Which one happens is decided by which member is weaker.
On a cast base the fastener is bought to a declared property class; the internal thread is produced in whatever the casting happens to be, by whatever tool was in the machine. The weaker member can be the one nobody specified to a strength.
Stripping explains the second base. It is gradual: it starts at the first engaged turn and works along the helix, and needs something like half a thread pitch of displacement before it is complete — so a joint can be partially stripped, assemble, torque up and go into service.
Read that as a production problem: a partially stripped joint can pass the assembly station, a leak test and a sea trial. The metal that was carrying the load is gone, and the only sign is a slightly longer turn than the one before it — the difference a busy line does not record. Weeks later it surfaces as a loose bolt. Losing clamp load is the episode on why marine hardware bolts loosen; this is about the material that should have held the clamp.
Engagement Depth Is a Material Question
Every drawing answers, by commission or omission, how much thread the joint may rely on. The answer depends, and one published analysis shows how far: an M12 class 10.9 fastener with 15 mm of engagement was assessed as adequate in spheroidal graphite iron and insufficient in aluminium, where the same analysis called for a minimum of 25 mm. Same bolt, same length, different answer — because the parent metal changed.
One bolt, one bracket, one analysis: an illustration, not a rule. The parent metal decides whether the fastener or the hole fails first. A strong one carries enough shear that the bolt limits; a weaker one strips its own thread while the bolt is intact.
Material around the hole is working too: the boss expands and the shear plane moves outward, and published analysis puts the useful limit for that stiffening at about twice the thread diameter.
So required engaged length is an output of the joint — parent metal, thread form, class, boss geometry, fastener strength — not a constant to look up. A universal minimum is either a guess or a summary of somebody else's joint.
The Hole Behind the Thread
A drilled hole is not a cylinder: it is usually a little wider at the entry than deeper in, over something like half a diameter, because the drill flexes as it starts. Tap it and the thread at the top is shallower than below. Whether the drawing called for a drilled or a cored hole is a starting point it should have named.
On a drilled hole the first engaged turn carries most of the load and is also the turn with the least thread height. A long engagement dilutes that; a short one does not, and with a fine pitch the loss is a bigger share of a smaller thread.
Blind holes add specification requirements, not drilling preferences. Drilled depth must exceed threaded depth, because a tap needs somewhere to start and somewhere to finish. Give a thread depth and no drill depth and the shop invents the difference — where a tap breaks, or the last full thread ends too high.
Chips and coolant then settle in the bottom, and an upward-facing blind hole collects water — a corrosion point covered in the episode on why 316 stainless steel rusts. A wet hole and a dry hole are different joints.
Chasing a higher percentage of thread height is not free: once strength is adequate, the extra percentage mainly raises tapping torque and tap breakage.
What a Class Does and Does Not Say
A class is easy to misread on a thread drawing. A class such as 6H is a tolerance on the thread, expressed as a grade plus a position. It constrains how the flanks may sit; it says nothing about how much load the hole carries.
A 6H hole is one whose flanks fall inside a band; it is not certified to hold a stated load, and another class is not a stronger joint. The class controls fit and interchangeability; strength is set by material and geometry.
Nor does it replace inspection. A thread that has to be released rather than admired is checked with a gauge, not by eye. The tolerance system behind that designation is ISO 965-1; the edition consulted for this article is the 2013 text. How a drawing fixes position across a family of fittings is treated in the episode on custom marine hardware.
Reading a Stripped Thread Without Overreading It
A stripped thread invites a story. Resist it and work the sequence.
- Which member lost material: bolt thread, hole thread, or neither. A fractured fastener is a different failure.
- Whether the fastener bottomed. A stud that ran out of thread before the joint closed shows it there, and shifts the diagnosis to wrong depth.
- Whether the joint ever reached clamp. If the parts never came together, the thread was loaded by assembly effort, not by service.
- Whether the entry was clean and chamfered. A burr means the first engaged turn was never a full one, and the turn carrying the most load started compromised.
- What the records say. Engagement as recorded, whether the hole was gauged, whether depth was checked, who signed it off — the only step free of interpretation.
The limit is the point: appearance cannot certify the strength that remains. A thread that looks intact after a partial strip may have lost the turns doing the work; one that looks damaged may have lost material that carried nothing. The bench can say which member gave way and whether the joint was built as drawn — not what it will hold now.
Repair is a change of design, not a restoration: the load moves into the insert and the thread that holds it, and the parent metal is whatever it was. Accepting one is a specification decision with a signature on it, not a workshop convenience.
What to Put on the Drawing and in the Goods-In Check
The drawing is where the shop's assumptions stop being assumptions.
| What the drawing states | What it controls | What it does not fix |
|---|---|---|
| Thread form, size and class | How the flanks are toleranced and how the thread is gauged | Whether the parent metal can carry the load |
| Required engaged length | How much thread the joint may rely on | A hole drilled too shallow or a boss too thin |
| Through or blind, with depth | Whether the tap has somewhere to lead and finish | Water and chips that stay in the bottom |
| Entry chamfer and hole condition | Whether the first engaged turn is a full one | Bellmouthing produced by the drilling itself |
Record the engaged length achieved, not the length called for, because the two diverge when it matters. Record the gauging result and the instrument, so the check can be repeated. Record who released the part: a record without an accountable signature is a file, not evidence.
None of this makes a thread stronger. It makes the joint's assumptions visible, so the next stripped thread reads back to a cause.
