
A stainless steel cylinder shell with top bore is a drawn cylindrical cup with an opening through the closed end, so the element or lead passes out through the top rather than the side. The bore through the end is the sealing and locating feature, so its diameter, position and edge condition are the real specification. The shell is otherwise a plain drawn cup, and the value is in holding the end flat and the bore true.
Blanking, drawing and piercing in sequence, then machining of sealing faces, at +/-0.01 mm on turned or milled features and +/-0.03 mm on formed details, with dimensional reports accompanying the first article. So the tight band is on the machined features – which on this part is essentially the thread and whatever sealing face exists – and the tube and the tip are at a band that reflects the drawing process rather than the machining.
The thread is the mount and the seal, and those are different jobs
- A tapered thread on a thermowell seals by interference. The flanks wedge, the metal deforms slightly, and the seal is distributed along the engaged length. It is reliable, it needs no gasket and no sealant, and it is the reason a thermowell can be screwed into an existing boss.
- And the thread also carries the mechanical load of the probe, which on a thermowell in a vibrating vessel is not nothing. So the thread is a structural member as well as a seal, and its engagement and its minor diameter are both structural dimensions.
- Which means the failure is not “it leaks” – it is “the thread has stripped and the probe is loose”. And a stripped thread in a pressure application is a serious outcome, because the probe may be the only thing keeping the process contained.
So the thread on this part needs three specifications and drawings usually give one: the standard and class, the sealing method, and the installation torque. The torque is a design output rather than an instruction because the torque is what makes the interference seal, and a probe installed at the wrong torque will either not seal or will be over-stressed.
The insertion depth, which is the accuracy
This is the point that appears on every thermowell page in this section and it is worth stating here in its simplest form, because on a cylinder-tube thermowell there is nothing to hide behind.
- Heat conducts from the process along the tube toward the mount, and the junction sits somewhere in that gradient. So the reading is displaced toward the cold end, and the displacement scales with how much of the tube is outside the process.
- Which means the number that matters is not the probe’s length – it is the sealed length, the part that is not in the process. And that number is not a property of the probe. It is a property of the boss’s depth and of whether a gasket or a seal is used.
- So the drawing has to carry it as a stated requirement even though nothing on the part produces it. Where it is missing, the probe is made exactly to drawing and reads several degrees out, and no inspection on the part will find it – because the part is correct.
The tip, and the thinning that the press decided
- A drawn closed end is thinned at the punch nose, so the tip is the thinnest section of the pressure boundary. On a low-pressure thermowell that is entirely acceptable; on a high-pressure one it is not, and the answer is a heavier gauge with a welded or machined tip.
- So the tip’s minimum wall is a specification, checked at the tip, and it is a different number from the blank’s thickness.
- And the tip’s thickness is also an accuracy parameter – a heavier tip conducts worse. So the same dimension is a pressure requirement and a response requirement, and where the two conflict the resolution is a heavier gauge rather than a thicker tip.
- And the internal surface finish at the tip, where a contact sensor sits. On a thermocouple in a thermowell the junction sits at the tip and the tip’s internal finish affects both the contact and the thermal path.
Why cylindrical, and why it matters
The word “cylinder” in this part’s description is doing real work, and it is worth understanding what the shape buys and what it costs.
- A cylindrical tube draws very cleanly – there is no corner to form, so the whole section is subject to the same springback and the wall is uniform. That is the shape’s main manufacturing advantage and it is a genuine one.
- And it is the right shape thermally. A cylindrical tip has no flat and no corners, so there is no local heat concentration and no local thinning. Its surface area for a given volume is the lowest of the practical tip shapes, which makes it a slow but even sensor.
- Where it does not help is where the tip needs to contact something – a surface temperature measurement needs a flat or a shaped tip, and a cylindrical tip pressed against a surface will only touch on a line and will read the contact’s own temperature rather than the surface’s.
So the shape answers one question: is this probe measuring a fluid or a surface? If a fluid, a cylindrical tip is the correct and simple answer. If a surface, this shape is the wrong choice and the design needs a different tip entirely.
Where the tip is welded, the L grade is not optional
Standard stock: carbon steel, 304 and 316 stainless steel, aluminium and brass, with grade and temper confirmed in writing before tooling. Three points.
- 316L for a chloride-bearing process and for a cleaning regime, and the L grade where the tip is welded.
- And the thermal conductivity, which on this part points the opposite way from intuition. Stainless Steel conducts poorly, and on an insertion thermocouple that is an advantage: a poor conductor flattens the temperature gradient along the tube and reduces the conduction error. So on a thermocouple thermowell, the material’s poor conductivity is doing useful work, and choosing a more conductive alloy to “improve” it would make the probe worse. On a resistance thermometer in the same housing, the same property is a disadvantage.
- Carbon steel where the process is benign and the temperature is moderate, with a coating. On this part the coating has one specific effect worth naming: a coating on the process-facing tip is in the thermal path and changes the sensor’s response, and on a contact-temperature application that is a measurement change rather than a finish.
On a thread that seals by interference, coating is the wrong answer
Finish selected from the drawing, with masking and laser marking available. On a cylinder-tube thermowell the finish’s requirements are few and specific.
- The thread, masked or specified post-finish with an allowance – on a thread that seals by interference, a plating changes the interference and it flakes at the roots, and the flakes go into the process.
- The tip, unmasked, where it is a contact surface or a product-contact surface.
- And on stainless steel, passivation rather than plating for the same reasons as everywhere else on this site – and on this part specifically, because a thread that seals by interference cannot tolerate a coating between the flanks.
The tight band goes to the thread, with the installation torque stated
+/-0.01 mm on turned or milled features and +/-0.03 mm on formed details. On this part the flagged dimensions are four, and the tight band correctly goes to the thread.
- Thread: standard, class, pitch diameter and taper angle, with the installation torque stated. This is the tight band and it is where it belongs.
- Any sealing face, flatness and perpendicularity as two values – relevant only if the design uses a face seal rather than a taper.
- Bore concentricity to the thread axis, where the tip’s position matters, because a thermowell whose axis is eccentric in its boss reads differently at different depths.
- Length, as a functional dimension with a tolerance, because it is the insertion depth – though the number that actually matters, the sealed length, is not a dimension of the part at all.
- And the requirements that are not dimensions: the insertion depth, the sealed length, the tip’s minimum wall, whether the arrangement is grounded, and the thread’s sealing method. Those five are the specification, and on this simple a part they are the whole content of the drawing note.
On a plain sleeve the assembly is the only seal verification
Dimensional reports accompanying the first article. On a part this simple the inspection plan has two items that are not standard.
- A functional leak or pressure test on the assembled probe in a representative boss, because on this part the seal is made by assembly and only assembly verifies it.
- A functional temperature check where an element is fitted, because the probe’s own accuracy depends on the insertion depth at use and not on anything that can be measured on the part.
- Thread gauging on every part, and the tip’s minimum wall on a first article.
- And a stated note on what cannot be checked: the accuracy claim depends on how it is installed, and the drawing should say so rather than leaving the number unqualified.
How deep does a long thermowell actually sit in the process?
How deep will this thermowell actually sit in the process, and what is the boss it goes into? Everything about this part’s accuracy is decided by those two answers and neither is a property of the probe. A 150 mm thermowell that sits 40 mm into the process is a different instrument from the same thermowell at full immersion, and the difference can exceed the entire accuracy claim. And the boss matters as much as the depth – a boss with a 20 mm internal thread leaves 130 mm of probe hanging in the air conducting heat from the room, and a boss with a 150 mm thread leaves nothing. So the honest question is not “what is the probe’s length” but “what is the immersion length and what is the sealed length, and who controls them”. Where the customer controls them, the drawing has to state them as requirements. Where the probe’s supplier controls them – which is the case when the probe and the boss are designed together – then they belong in the design and not on the probe’s drawing. And the second question, on a simple part like this one, is worth asking anyway: is a threaded thermowell the right part at all, or should this be a flanged or a welded probe? A threaded thermowell is the cheapest to install and the easiest to replace, and it seals reliably. It is also not removable once it has been in service, because the thread has deformed. On a vessel that has to be inspected internally, that is sometimes the wrong trade and it is worth naming at enquiry rather than at the inspection.
Parts like this one sit under the Sensor Housings range, whose overview page gathers the tooling logic, finishing options and tolerance notes shared across the family.
Specifications
| Item | Stainless Steel Cylinder Shell with Top Bore |
|---|---|
| Brand | AUNOV |
| Designing | Made to your 2D/3D drawings (PDF, DWG, STEP, IGS) – unusual walls, threads or tolerances are feasibility-checked first; the agreed version governs every later batch. |
| Tooling | Dies, fixtures and gauges are made in our own toolroom, so die cost stays transparent, and ownership is confirmed on the quotation. |
| Samples | Sample parts are measured and reported before submission, so the approved first article becomes the gauge reference, and later batches are gauged back to the signed sample. |
| Materials | Standard stock: carbon steel (SPCC / DC01), 304 / 316 stainless steel, aluminium, brass; grade and temper are confirmed in writing before tooling. |
| Surface finish | Plating, coating, anodising or passivation are chosen from the drawing; masking and laser marking are available where the job needs them. |
| Processing | Blanking, drawing and piercing in sequence, then machining of sealing faces; every stage is logged against your part number. |
| Equipment | Press line for drawing, CNC cells for finishing, with CNC turning centres for secondary operations and CMM inspection. |
| Tolerance | +/-0.01 mm on turned or milled features, +/-0.03 mm on formed details; dimensional reports accompany the first article. |
| Delivery terms | Incoterms FOB as standard with DDP on request; sea and air options priced with the order. |
Related Products
Frequently Asked Questions
Why is a cylinder-tube thermowell harder than it looks?
It is the simplest thermowell there is: a tube with a thread on the outside and a closed tip, screwed into a boss, with no gland, no nut and no fillet. Because it is that simple, all the difficulty has moved into three places easy to leave out of a drawing, the insertion depth, the tip’s minimum wall, and the fact that the thread is both the mount and the seal. Blanking, drawing and piercing in sequence, then machining of sealing faces, runs at +/-0.01 mm on turned or milled features and +/-0.03 mm on formed details.
What three specifications does this thread actually need?
The thread on this part needs three specifications and drawings usually give one: the standard and class, the sealing method, and the installation torque. On a thread that seals by interference, coating is the wrong answer because it changes the fit. On a plain sleeve the assembly is the only seal verification, so the torque has to be stated with the part.
How deep does a long thermowell actually sit in the process?
A 150 mm thermowell that sits 40 mm into the process is a different instrument from the same thermowell at full immersion, and the difference can exceed the entire accuracy claim. The boss matters as much as the depth: a boss with a 20 mm internal thread leaves 130 mm of probe conducting heat from the room. Standard stock is carbon steel, 304 and 316 stainless steel, aluminium and brass, with grade and temper confirmed in writing before tooling.
Where the tip is welded, why is the L grade not optional?
Where the tip is welded, the L grade is not optional because the welding heat would otherwise sensitise a standard grade at the thinnest, hottest section. On a cylinder-tube thermowell there is nothing to hide behind, so this point appears here in its simplest form. The flagged dimensions are four, and the tight band correctly goes to the thread.
Request a Quote for This Part
Send a drawing or sketch — get material advice, tooling cost and a per-piece price within 24 hours (working days).




