Exploring the components of On-Machine-Inspection

Topic:

Why probe on your machines?

Why should you want to measure parts on the machine, and can it be done?  Won't using the machining to inspect take away from cycle-time?. This may sound counter-intuitive, but not when you define cycle time correctly. Cycle time is not simply the time spent machining. True cycle time is defined as the rate at which you make and certify good parts:

Cycle time = {Set-up time + machining time + certification time} / number of good parts produced

Probing on the machine shortens cycle time by reducing:

• Set-up time - probing can automate part and tool setting, reducing downtime.

• Certification time - post-process, off-line inspection is often undesirable and, in some shops, impossible: queue and move times and the cost of capital can be prohibitive. Manual methods need extra gauges and are prone to human error. By contrast, machine tool probes can measure the part without delay and without additional equipment or labour.

• Scrap - automated, precision feedback at both the setting and inspection stages can eliminate operator errors and compensate for process drift.

Topic:

How can you trust the machine to measure its own work?

How do you trust your micrometer to be accurate... Establish TRACEABILITY - i.e. a quantified level of uncertainty against national and international measurement standards.

• Establish and PROVE the measurement capability of your machine tools.

• Quantify the measurement performance of your machine tools and introduce traceability to the inspection process.

• By using certain metrology techniques, you can identify and minimise measurement errors on the machine tool, allowing you to use it to control its own production.

What do I need to do?

• Use Renishaw inspection probes and tool setters to monitor and control the machining process.  Introduce traceability through artifact comparison, footprinting or through derived measurement uncertainty

Topic:

Artifact - what are they and how are they used?

Introducing traceability using comparative measurement techniques

What is an artifact? ...a workpiece that has been measured at 68F on a calibrated CMM, but which lives in the machine, made from the same material as components to be machined, ideally positioned in the same area of the machine as the workpiece.

What types of artifact are there?

• generic artifacts - collection of standard features

• replica components - exact match with parts being machined

How does artifact comparison work?

• artifact dimensions measured on calibrated CMM

• calibrated dimensions stored in CNC

• artifact probed on the machine tool, and actual dimensions are computed - includes growth and distortion due to thermal effects, plus systematic errors in the machine structure

• difference between actual and calibrated size is stored as a correction factor - either a specific correction for each feature, or a scaling factor

• component is measured, and correction factors are applied to remove the current machine errors detected by measuring the artifact

The ideal artifact comparison process!

Factors that affect the level of measurement uncertainty using artifact comparison techniques:

• artifact stored in machine all the time and flooded with coolant

• same temperature as the workpiece

• artifact made of same material as the part

• same expansion coefficient

• artifact measured in same part of the machine tool

• same local machine growth and distortions

• artifact and workpiece measured with minimum time delay

• no transient changes in temperature can have a significant impact on the results

Topic:

Artifact comparison - how accurate is it?  What levels of traceable measurement error can be achieved on machine tools?

Quantifying the measurement uncertainty

Establish the traceable accuracy of your machine tool probing by quantifying difference between machine tool & CMM results and adding it to the known CMM error.

Measurement uncertainty to NPL / NIST / PTB = [(CMM error)2 + (MT diff)2 + (Probe cal error)2]1/2

Our own results - traceable errors:

Our own results - traceable errors:

On our own machine tools, using various types of artefact, we have demonstrated traceable measurement errors as follows: dedicated artefact (replica component, feature-specific corrections):

• ±0.006 mm (±0.00024 in) error (95% confidence)

generic artefact (scaling factor correction):

• ±0.012 mm (±0.00048 in) error (95% confidence)

Artifact comparison - step by step.

How we use artifacts in its own machine shop.

1. Machining

• component machining with coolant on

• artifact experiences same ambient conditions as workpiece

• short coolant wash after machining complete to generate uniform temperature and remove residual heat from workpiece

2. Probe artifact

• artifact indexed into probing position

• measurements taken and correction factors computed

3. Probe components

• immediately index first component fixture into position

• measure component features and apply artifact-derived correction factors

• take action on the results:
• accept / reject part
• update offsets

Topic:

Errors that prevent traceability if not corrected

Transient and permanent error sources

There are two main sorts of error on machine tools that must be quantified or eliminated if traceable measurement is to be achieved:

• Thermal growth and distortion - a transient / changing phenomenon.

machine tools are generally not in temperature controlled environments...

ambient temperature affects the part and the machine

temperature gradients can cause distortion in the part / machine

machine tools generate heat...

heat soak from the spindle motor can affect Z axis

axis drive motors and friction in bearings can affect ball screw length

• Machine geometric errors - a consistent / underlying error source

some errors can be compensated in the CNC, but others remain...

squareness, straightness, angular

Topic:

How do you know your parts are correct?

Inspect parts on your machines to a traceable standard - i.e. where you know the uncertainty of measurement

So, your machines are now performing to the best of their ability, but you still face the day-to-day challenge of getting good parts from them, in the face of variations in tooling, materials, the shop environment and operator error. The challenge is this: prove whether or not your parts meet specification whilst they are still on the machine. But can you trust the machine that made the parts to check them using touch probes?

How do I accept parts off my machine as being correct?

Monitor, diagnose and improve the performance of your machines, and use traceable in-process feedback to control your machining

Accepting parts off your machines - using the machine tool to control and monitor its own output - is a rapidly growing trend as manufacturers strive to make:

• 100% good parts, right first time

• in the lowest possible cycle time

• using established technology

Reducing cycle times whilst buying parts off your machines means meeting the following challenges head on:

• Make parts right first time by making sure your machines are capable of producing parts to specifications - before you start cutting. If the machine is correct, the parts coming off the machine should also be correct, barring material imperfections, catastrophic failures of tooling & fixtures, or a machine crash.

• Prove that your parts meet specification by introducing process control to traceable standards on the machine. If you know the measurement uncertainty on the machine, you can use it to control your process and to pass off parts.

Topic:

How do you make parts right first time?

• Identifying and improving the capability of your machines

• PROVE the capability of the machining process

• Quantify and, where sensible, reduce machining errors, and apply a process control regime on the machine that ensures that the design tolerances can be met.

• measure to a recognized standard, such as ASME B5.54, B5.57 and ISO 230

• provide certified documentation demonstrating the capability of the machine tool and control of the process

How? - use the ML10 Gold Standard laser calibration system and the QC10 ballbar to measure the capability of your machines

Topic:

In-cycle probing for traceable process control

Establishing a feedback regime that takes account of your machining and measurement process capabilities

If you make parts in batches or longer production runs, you will be concerned about controlling process variation. Now that you have optimised your machine performance and established a traceable inspection process on your machines, your final step is to establish an appropriate inspection and process control regime. Here’s how:

Establish machining process capability

• Understand the inherent variation and underlying rate of drift in your machining processes by measuring parts either on the machine, using one of the traceable techniques outlined above, or on your traceable CMM. Compare this capability with your part tolerances to determine how often you will need to inspect the parts to keep the process under control.

• understand tool wear, thermal drift, fixturing repeatability

• compare with drawing tolerances to determine inspection frequency

Account for inspection process capability
Know the traceable accuracy of your on-machine measurement, including measurement repeatability. Take account of this when setting your process control thresholds.

• establish repeatability by measuring a feature 10 times

• establish accuracy by comparing machine tool probing results with measurements on your CMM

• take account of measurement uncertainty when setting process control thresholds

Set up a process feedback regime
You don’t want to probe more features than you have to, so identify those that are critical to product performance. Ideally, measure one control feature per finishing tool, and use the measurement error to update the size or length offset for that tool.

• inspect critical features - at least one per finishing tool

• update tool offsets

• apply damping factor to feedback

• set alarms if results fall outside control thresholds

• log results for SPC

• check for broken tools to prevent scrap

Probing software, makes process control simple, by setting control limits and applying a damping factor to offset corrections. You can, of course, also use tool setting to help make parts right first time by ensuring that tool offsets are correctly set at the start of a production run, and include broken tool detection to increase your confidence in unmanned machining.

Overview: Can your machines be used to measure parts?

CMM @ 20°C (68°F)

• provides the ‘final say’ on workpiece acceptance

• calibrated with a laser

What about machine tools?

• CMM and machine tool structures are very similar

• both calibrated to national / international standards

• CMM can provide periodic traceable check on machine tool probing

Return to ProbeDriver Main Page

 Images and a portion of text used, by permision, from Renishaw PLC all rights reserved.