Stainless Steel Stepped Sensor Sleeve with Base Flange

Stainless Steel Stepped Sensor Sleeve with Base Flange
Stainless Steel Stepped Sensor Sleeve with Base Flange

A drawn temperature sensor housing that is finished on a CNC is the same product as the previous page and a different engineering problem, because the emphasis has moved. Where that page was about the split between drawn and machined tolerances, this one is about the transition itself – the corner where the press stopped and the lathe started – and about the specific defects that live in that corner.

The corner is where a drawn-and-machined part fails, and the failures there are not intuitive: they are not the failures anyone checks for.

The transition, and its three specific problems

A drawn shell has a corner where the wall meets the closed end – the bend radius. And it has an open end that the machining has to work from. The transition between the drawn features and the machined features is a real geometric discontinuity with three practical consequences.

  1. The corner’s work-hardened state. The bend radius is the most heavily worked metal on the entire part. It has been stretched on the outside and compressed on the inside, and it has hardened. On a thin-wall drawn part this means the corner behaves differently from the rest of the wall under any subsequent forming, handling or thermal cycling. It also means a fixture that clamps on the corner is clamping on hard, inconsistent material.
  2. The residual stress in the corner. Drawing leaves residual stress and the corner carries the most of it. Under thermal cycling – and a temperature sensor housing is specifically a part that experiences temperature cycling – the corner is where that stress is released first. On a thin drawn wall this is where a crack initiates.
  3. The mismatch between the drawn dimension and the machined one. The open diameter comes from the die, the machined features come from the tool, and the two are made in different operations with the part held differently. Their relationship is therefore a tolerance and not a consequence – and that tolerance is the one that has to be on the drawing.

Concentricity, which is the tolerance this part is really about

On a temperature sensor housing, concentricity is functional rather than aesthetic, for a reason that is easy to state and easy to forget.

  • The sensing element has to be in the right place in the housing’s interior. The element sits against the closed end, or against a thermal mass, or inside a bore. If the machined seat is not concentric to the drawn bore, the element sits off-centre, the thermal path becomes asymmetric, and the sensor’s reading becomes dependent on its own orientation.
  • On a thermowell-style design, this is exactly the same subject as covered on the thermowell page: the accuracy is set by the geometry of the interior and by the insertion depth. The difference here is that the interior’s geometry came from a press rather than from turning, so the accuracy comes from the blank’s thickness and the draw rather than from the machine.
  • So the specific tolerance to write is: the concentricity of the machined seat to the drawn bore’s axis, held independently of the seat’s size, and with the datum being the drawn axis rather than the machine’s origin. And where the sensing element sits on a thermal path, the same concentricity is a measurement-accuracy requirement and not only a fit requirement.

Where the press stops, which is where the designer has to think

Which features should be drawn and which should be machined is a design decision, and there are three principles for making it.

  • Draw everything that does not need a tolerance. The wall, the open diameter, the corner radius, the general shape – all of these are cheaper to draw and the drawing process makes them consistent within the material’s own variation. Machining them would be an enormous cost for no benefit.
  • Machine everything that seals, locates, or threads. Sealing faces, spigots, threads, and any seat that has to be concentric to something. These are the features where a few hundredths of a millimetre matters and where the press cannot deliver it.
  • Design the joint between the two. This is the part that is genuinely hard, and it is worth being specific about what makes it hard. The machined feature has to be cut on a surface the part can be held against, and that surface is a drawn one – so its flatness and its position are set by the press and the machining has to work from them. The best answer is to put a flat or a step in the drawn blank specifically so that the machining has a repeatable datum to work from, and to designate that flat as the housing’s single machining datum.

Wall thickness and thinning, revisited with the machining in mind

The same two mechanisms from the previous page, but with a specific new consequence because the part is machined afterwards.

  • The punch nose thinning at the closed end is the housing’s minimum wall. On a temperature sensor housing the closed end is usually where the thermal path is most sensitive, so the thinning is a specification. And it should be stated as a minimum at that location, checked there, because it is the one place on the part where the wall is thinner than the blank.
  • Machining near the thinned region is risky. If a machined feature is cut close to the closed end – a small bore, a counterbore, a thread – the tool is cutting into metal that has already been stretched and has less thickness than the drawing suggests. The wall’s remaining thickness after machining has to be checked, and on a thin housing this is a real design constraint rather than a detail.
  • The corner radius and the wall. A deep-drawn corner radius on a thin wall means the effective wall at the corner is the thin section at the bend’s inside. On a housing where the corner is where the element or the seal is, the corner’s inside thickness is the number that matters, and it is smaller than the wall’s nominal.

A sleeve’s grade follows the cleaning regime as much as the chloride count

Carbon steel, 304 and 316 stainless steel, aluminium and brass are all held on site, with material substitutes proposed only with the written approval. Four points on this part.

  • 316L for the general case. On the chloride logic and the cleaning-regime logic that applies to any wetted stainless steel part, and the L grade where there is welding or heat input.
  • Where a temperature sensor is measuring a fast transient or a small feature, the housing’s thermal conductivity starts to matter and aluminium becomes a legitimate answer. The consequence is the one that is easy to miss: aluminium expands about twice as fast as the steel leads and terminals the sensor will be connected to, and a hermetically sealed aluminium housing with a steel terminal will see differential movement at temperature. That is a real design consideration and it belongs at the enquiry.
  • Where the housing will be welded, the grade’s weldability is a requirement and 316L is the safe answer. A welded austenitic stainless steel housing that has been sensitised will corrode in the heat-affected zone, and that corrosion is inside the part where nobody will look at it.
  • Brass appears where thermal conductivity and non-sparking matter, and it is easy to machine. Its limit in a sensor housing is dezincification and the fact that it is a poor choice where the process is a strong alkali or a strong acid.

A welded housing needs its own treatment spec for the weld region

Plating, coating or polishing to suit the drawing, with masking and laser marking available. Three items.

  • The sealing face unmasked, as always, and where the housing is welded the weld region needs its own treatment specification rather than a general one.
  • The thread masked or specified post-finish with an allowance.
  • Passivation on stainless steel, which is required after both the drawing and the machining, and which should be recorded as an operation rather than assumed.
  • And where the housing is a visible part – on an instrument, or a panel-mount probe – the external finish becomes an appearance requirement and the approved sample is the specification. On a small drawn part with a machined band, it is worth noting that the drawn surface and the machined surface will take a finish differently, and that the transition between them is visible. This is a real appearance consideration and it is worth deciding the geometry with it in mind.

The machined seat’s concentricity to the drawn bore is its own characteristic

+/-0.01 mm on sealing faces, +/-0.05 mm on drawn profiles, with finer bands confirmed from capability data before quoting. The split is right, and three specific tolerances complete it.

  • Concentricity of the machined seat to the drawn bore, as its own characteristic, with the drawn axis as the datum.
  • The minimum wall at the closed end, measured after forming, stated as a minimum rather than as a nominal.
  • The machined features’ position along the axis from a named datum that is a drawn feature, so that the machining’s position reference is the press’s output rather than the machine’s arbitrary face.

The general principle worth carrying: on a drawn-and-machined part, every machined dimension should be referenced to a drawn datum. Because the drawn features are what the material decides and the machined features are what the machine decides, and the only stable relationship between them is the one the drawing states.

Wall thickness at the closed end needs its own number, not the side value

  1. Sealing face flatness and perpendicularity, on every part.
  2. Wall thickness at the side and at the closed end, measured separately, with the closed end compared to its minimum.
  3. Concentricity of the machined features to the drawn axis, on the CMM. This is the inspection that is specific to this part.
  4. Thread gauging and, where the housing has a wire or terminal, a check of the wire’s seal and the terminal’s fit.
  5. A functional leak or hermeticity test, on every part where the sensor’s specification requires it.
  6. First articles and batch samples on the CMM, with in-process checks between the drawing and the machining stages rather than only at the end.

Whether the part can be drawn with no machining at all

Can the whole part be drawn – no machining at all? This sounds obvious and the answer is often yes. A drawn sensor housing with a drawn open end, a drawn spigot and a drawn or rolled thread is entirely feasible on many designs, and it is dramatically cheaper and dramatically more consistent than a drawn-and-machined one. The reasons a drawing gets machined are usually that the design needs a sealing face flatter than the press can hold, or a thread that has to be a specific standard and fine pitch, or a seat that has to be concentric to the bore within a band the press cannot deliver. Each of those is a real reason, but each of them should be a decision rather than a default. So the useful question is: which of the three reasons applies here, and can the design be changed so that it is the first or the second rather than the third? A housing that needs one machined flat and one machined thread is a normal part. A housing that needs four machined features because the design was drawn without thinking about which features the press could make is a design that costs more than it needs to. The answer decides the process route, the cost and the consistency, and it is worth asking before the drawing is complete rather than after the first quotation.

You will find this part in our Sensor Housings range. Process notes, common materials and inspection points for every product in the family are collected across the sensor housings pages.

Specifications

ItemStainless Steel Stepped Sensor Sleeve with Base Flange
BrandAUNOV
DesigningQuotation follows the print, not a catalogue – unusual walls, threads or tolerances are feasibility-checked first; the agreed version governs every later batch.
ToolingOur toolroom covers die design, wire-cut EDM and maintenance, so a die can be re-cut against the approved drawing, and maintenance is tracked against the part number.
SamplesSample parts are measured and reported before submission, so any deviation is corrected at sample stage.
MaterialsCarbon steel, 304 / 316 stainless steel, aluminium and brass are all held in stock; material substitutes are proposed only with your written approval.
Surface finishPlating, coating or polishing to suit the drawing; masking and laser marking available where needed.
ProcessingDeep drawing on hydraulic presses, CNC-finished where tolerances tighten; in-process checks sit between stages, not only at the end.
EquipmentPress line for drawing, CNC cells for finishing; first articles and batch samples are checked on the CMM.
Tolerance+/-0.01 mm on sealing faces, +/-0.05 mm on drawn profiles; finer bands confirmed from capability data before quoting.
Delivery termsFOB Ningbo or Shanghai, payment by T/T, L/C at sight or D/P; repeat orders reuse the approved first article.

Frequently Asked Questions

Why is the transition between drawn and machined the real risk on this sleeve?

The corner where the press stopped and the lathe started carries the most work-hardening and the most residual stress, and under temperature cycling that is where a crack starts. The mismatch between the drawn diameter and the machined one is a tolerance, not a consequence, and it belongs on the drawing as its own callout.

Why is concentricity the tolerance that matters most here?

If the machined seat is off the drawn bore’s axis, the element sits off-centre and the thermal path becomes asymmetric, so the reading depends on the housing orientation. The specific callout is the concentricity of the machined seat to the drawn axis, with the drawn axis as the datum.

Why does the grade follow the cleaning regime as much as the chloride count?

On a food, pharmaceutical or medical part the cleaning agents are often more aggressive than the process fluid, so the grade is chosen for the wash as much as the fluid. Where the housing is welded, the L grade also resists sensitisation in the welded heat-affected zone, which would otherwise corrode where nobody looks.

Can the whole sleeve be drawn with no machining at all?

Often yes – a drawn open end, spigot and even a rolled thread is feasible and a drawn and rolled sleeve is cheaper and more consistent than a machined one. Machining is added only when a face must be flatter than the press holds, a thread must be a fine standard pitch, or a seat must be concentric within a band the press cannot reach, and each should be a decision rather than a default.

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).