There is a titanium pylon bracket waiting to be programmed. An 860mm Ti-6Al-4V forging, destined for a large horizontal machining centre. It is the kind of part where the stock is expensive, the lead time is long, and the first cut is effectively the only cut. Get it wrong, and the cost is not an hour of rework. It is a scrapped forging and a schedule that slips.

This is the part on the programmer’s screen this morning. Here is how the day goes when DigitalCNC sits alongside the CAM system.

The program that looks finished

By mid morning, the toolpaths are done. Rib top roughing, two levels of roughing, finishing passes on the open faces and the narrow features. In CAM it all looks right: the strategies are sound, the engagement is controlled, the simulation runs clean and the part comes out to shape. On any normal day this is where the program is posted, taken to the machine, and proven out, with the programmer standing at the control, listening to the cut, watching for stutter and dwell, making changes on the floor while the spindle runs and stock is consumed.

Instead, the posted program goes through DigitalCNC first. It does not connect to the machine and it does not read anything off the floor. It runs the posted code through a model of this specific machine and controller, how the look-ahead reads the program, how the axes accelerate and decelerate, where the kinematics constrain the motion, and predicts how that machine will actually drive the toolpath. At the desk, before any metal is touched.

A word on the one number that recurs below. Efficiency here is not a marketing figure. It is how much of the commanded motion the machine will actually realise: the ratio of the feed the machine can hold to the feed the program asks for, along the path. Ninety per cent means the machine delivers nine tenths of what the toolpath assumes. The gap is time the programmer is paying for and cannot see in CAM.

Titanium Pylon Bracket Forging Model in Mastercam

Decision one: the corner the machine cannot turn

The first thing the analysis surfaces is the rib top roughing. The program commands 2,200 mm/min. The prediction shows that through the corners the machine will only ever deliver 500. The toolpath could not show this, because the toolpath does not know the machine. It would have been the first thing the programmer heard on the floor: the feed dropping away, the cut labouring through every corner.

Now it is a number on the screen instead of a sound on the machine. The fix is made in CAM: allow 5mm of corner rounding where the geometry permits, and the commanded feed returns to where the programmer intended it. One change, made upstream, on evidence rather than instinct.

Decision two: the finishing pass that was quietly underperforming

Next is the first finishing pass, running at 82 per cent: the machine is only achieving 20% of commanded feedrate. The analysis shows why. Chip thickness is collapsing at ramp in and ramp out, and on titanium that is the worst place for it, the regime where the tool starts rubbing rather than cutting. The time domain plot shows exactly where along the path it happens.

The programmer opens the tolerance from 25 to 80 microns in the regions where the geometry allows it, and re-analyses. Efficiency comes back to 89 per cent. What matters is not only the recovered time. It is that the decision to relax tolerance was made knowing precisely where it was safe and what it would buy, rather than applied as a blanket guess and discovered on a CMM days later.

Decision three: turning a guess into a defensible number

The second level of roughing sits at 90 per cent. Good, and the analysis shows the one thing holding it back: a 5,000 mm/min repositioning feed the programmer had entered as a reasonable estimate. The machine can only deliver 3,000 to 3,500.

That repositioning feed was a guess, the kind every programmer makes a dozen times a day because there is no better number to hand. Now it is replaced with what the machine will actually achieve. The guess becomes a figure the programmer can defend: to a customer, to an estimator, to themselves.

Decision four: protecting the tool on the cut that matters most

The hardest one is the finishing pass on the narrow features, at 32 per cent, hard against the acceleration limit of the platform. The obvious lever, relaxing tolerance further, is the wrong one here. It would put the tool tip at risk on a slender, vulnerable cut, late in the part, after hours of metal have already been removed. A tool failure or a defect there does not cost a pass. It costs the forging. So the decision goes the other way: bring the surface speed down on this final cut, where a mistake is most expensive and least recoverable, and protect the part the whole day has been building toward.

Not every problem has the same answer. The analysis gives the programmer the information to choose the right lever for each one, instead of applying the same fix everywhere and hoping.

The error you only catch by reading the code

There is one more, and it has nothing to do with feeds or tolerances. While the analysis runs, it flags feed call errors in the posted NC. Not in the toolpath, which was correct, but in the code the post produced from it. These are the errors that survive everything upstream, because the simulation ran the toolpath and the part came out to shape, and the fault is sitting in the posted output where nobody looks until someone reads the file line by line, or until the machine does something unexpected with an expensive forging clamped in it. Here they are caught and corrected before the program leaves the desk. It is the least glamorous part of the day and, on a part like this, quietly one of the most valuable.

The program that reaches the machine validated

By the end of the day the bracket has not been near the spindle, and that is the point. Every problem that would have been found on the machine has been found in CAM. Every fix that would have cost a prove out iteration has been made at the desk: the corner feed, the chip thinning, the repositioning feed, the acceleration limit, the posting errors, all resolved before the first cut.

When the program does reach the machine, the machine confirms it rather than discovers its faults. The forging is cut once, validated against the real machine rather than an assumed one.

What changed

Look at what the programmer did differently. They did not work harder or know more than they did yesterday. They made the same decisions they always make, on feed, tolerance, strategy and surface speed. The difference is that they made them against the real behaviour of the target machine instead of an idealised model, and they made them upstream, where a change costs minutes rather than metal.

That is the day. Same programmer, same part, same levers. A different place to find the problems, a different kind of confidence in the program that goes to the machine, and a forging that is cut once, and cut right.

The post A Day in the Life of a CAM Programmer Using DigitalCNC appeared first on DigitalCNC.

The post What’s New inside DigitalCNC (June 2026) appeared first on DigitalCNC.

AMRC / Tier 1 Aerospace Case Study – Available as PDF on our website

That gap does not appear in simulation. It appears during prove-out, on a machine that costs £5,000 a day to run. Working with the AMRC, we helped a Tier 1 aerospace supplier make that problem visible before the first cut. Multi-day iteration loops were replaced with analysis completed in under two hours, recovering up to £25,000 in machine capacity per cycle.

Read the full case study.

April was built around making that kind of analysis faster to access and easier to evaluate. Here is everything we delivered.

Virtual Environment: Try DigitalCNC in a Browser

Your CAM tells you what to cut. DigitalCNC tells you what will actually happen.

The virtual environment is now live. Step through real interactive use cases, select machines, compare strategies and quantify the gap in machine hours, cost per part and total programme impact. No download. No IT clearance. No installation. Register and you are in.

https://digitalcnc.ai/virtual-environment/

Native Integration: Mastercam

DigitalCNC now runs natively inside Mastercam. Siemens NX and CATIA V5 were already supported. Open it inside your CAM environment, select a toolpath, run the analysis. We are actively working on further integrations.

MACH 2026

We launched the virtual environment at MACH. The response was clear: teams want to evaluate toolpath performance against real machine behaviour without committing machine time to find out.

MACH also confirmed two things. Defence demand is drawing complex, tight-tolerance work into programmes that would not have carried that requirement five years ago. And simultaneous five-axis is now a production baseline, but the depth of process knowledge required to run it reliably has not kept pace with the installed base.

Read our key takeaways from MACH here.

ModuleWorks Insider Conference

We attended the ModuleWorks Insider Conference alongside Boeing, Sandvik, Dassault Systemes, Siemens, Mastercam, SolidCAM and a number of engineering-focused companies working on adjacent problems in CAD/CAM and CNC.

The clearest observation from the two days was that machining foundational models remain a long way from production use. That assessment came from the people working most closely on the problem, and it reflects why physics-informed, machine-specific analysis is the correct approach today rather than a position to revisit later.

Read more about what our CEO, Dr Rob Ward has to share about his experience at MIC 2026.

What Your CAM Software Does Not Tell You About 5-Axis Machining

The third session in our webinar series, delivered with experts from the AMRC, is now available in full on YouTube.

The session addresses feedrate fluctuations in simultaneous five-axis work, the conditions under which they produce witness marks and surface defects, and how to identify slowdown zones before the part is on the machine. It covers why three-plus-two is frequently the more reliable choice, how vector programming requires machine-specific validation and what offline testing looks like in practice.

Watch the recording on YouTube.

Cyber Essentials Plus Certification

If you handle sensitive geometry, proprietary process data or supply chain information under contract, the security posture of every tool in your workflow matters.

DigitalCNC has achieved Cyber Essentials Plus certification: independently verified against real-world attack scenarios. For teams operating in defence and aerospace programmes, this provides a recognised benchmark for data protection and supply chain integrity.

We are heading into May with significant partnerships and product announcements. Follow us on LinkedIn to see them first.

The post Monthly Rewind | April Newsletter appeared first on DigitalCNC.

WHY THIS PROBLEM EXISTS

When a CAM programmer sets a tolerance in their software, they are defining how closely the toolpath geometry should approximate the ideal part surface. What they are not defining is how the machine tool controller will interpret and execute that toolpath in practice.

The controller tolerance (CTOL) governs how the CNC interprets the stream of points and arc commands it receives. Where the CAM tolerance defines the shape of the toolpath on screen, the CTOL defines how faithfully the machine actually follows it. These two settings interact in ways that most CAM systems do not model, and that most programmers rarely have cause to examine directly.

The problem is compounded by machine kinematics. As the machine decelerates through tight corners, direction changes, or short segment sequences, the actual chip thickness delivered to the cutting edge drops significantly. In finishing operations, where tolerances are tight and surface quality is paramount, this deceleration can push chip thickness below the edge rounding of the tool itself, creating conditions for rubbing, work hardening, and possibly chatter rather than clean cutting.

CAM systems do not account for this by default because they have no understanding of the specific machine tool’s acceleration and deceleration behaviour through complex toolpaths. The toolpath looks correct in simulation because, geometrically, it is. The problem only becomes visible on the part.

WHY IT MATTERS IN PRACTICE

For general machining, a degree of surface variation is manageable. For aerospace components, it is not. Wing ribs, spars, and structural pockets are often manufactured to tight positional and surface tolerances, in materials that are sensitive to work hardening, and on features where rework is expensive or impossible.

The consequences of poorly understood tolerance interaction are not always dramatic. They often show up as: unexplained surface finish variation between nominally identical operations; chatter marks in deep pockets where tool extension is significant; premature tool wear that seems inconsistent with the programmed cutting conditions; and occasional scrap on high-value billets where the cause is difficult to diagnose confidently.

In deep pocketing conditions in particular, the compounding effects are severe. Cutter engagement spikes as the toolpath wraps around corners with no effective radial control. The machine decelerates through the corner. Chip thickness falls. With a long flute in contact, cutting forces rise sharply. In the worst cases, this produces chatter, surface defects, and spindle degradation on components where none of those outcomes are acceptable.

A 50% feed reduction in corners is a widely used CAM-level response to this problem. It helps, but as a blanket rule it does not account for the specific interaction of tool diameter, finishing stock, CAM tolerance output, and controller tolerance on a given machine. The result is a process that is partially compensated but not truly understood.

THE CONVENTIONAL APPROACH AND ITS LIMITS

The standard response to unpredictable surface finish in complex milling is conservative programming. Feed rates are reduced. Finishing stock is increased to give more margin. Corner feeds are halved as a rule of thumb. Trial cuts are run on representative material before committing to production.

Each of these measures costs something. Reduced feeds extend cycle times directly. Additional stock means additional finishing passes. Trial cuts consume machine hours and material. And even after all of this, the programmer is left with a process that has been de-risked rather than understood. If the surface finish is acceptable, they do not know how much margin they have. If it is not, they are adjusting parameters without a clear picture of which variable is responsible.

The deeper issue is that conservative programming optimises for avoiding failure rather than achieving the best possible outcome. On a high-value aerospace part, that distinction matters both in unit cost and in the confidence with which a process can be transferred, repeated, or quoted.

THE BETTER APPROACH

Tolerance-Aware Process Optimisation Using Real Machine Kinematics

The core of the approach is replacing assumed machine behaviour with modelled machine behaviour. Rather than accepting that the CAM simulation represents what will happen on the machine, the programmer works with a tool that understands the acceleration and deceleration behaviour of the specific machine, toolpath geometry and tolerance.

The starting point is the NC program as it will be sent to the machine. DigitalCNC ingests this directly and simulates execution against a model of the target machine tool, including its controller tolerance settings. The output is not a geometric verification but a kinematic one: what feedrate will the machine actually achieve at each point in the toolpath, and what does that mean for chip thickness at the cutting edge.

For finishing operations, the critical output is the feed per tooth map across the entire toolpath. Where chip thickness falls below the edge rounding value of the tool, the software identifies a risk of ploughing, work hardening, and chatter. The programmer can see exactly where on the part these conditions are likely to occur, and why.

This changes the decisions available to the programmer in several concrete ways. Tolerance settings, both CAM tolerance and CTOL, can be evaluated not just for their effect on cycle time but for their effect on cutting conditions. Corner strategies can be assessed against real deceleration behaviour rather than geometric approximations. Tool diameter selection for deep pockets can be informed by the actual wrap-around engagement the machine will deliver, not the idealised engagement the CAM system assumes.

The workflow change is significant but not disruptive. The programmer continues to work in their existing CAM environment. DigitalCNC operates as a plugin or parallel analysis tool, taking the NC output and returning a feedrate and chip thickness profile before the programme goes to the machine. Adjustments are made in CAM, reprocessed, and re-analysed until the achieved cutting conditions across the toolpath are within acceptable bounds.

What the programmer can now predict, and control is the actual chip thickness delivered at every point in the cut, on their specific machine, with their specific tolerance settings. That is the information that was previously only available after the cut had been made and the surface had been inspected.

PRACTICAL EXAMPLE

Deep pocket finishing in aluminium on a DST Ecospeed

Consider a deep pocket finishing operation on a structural aluminium component. The feature requires a long-reach tool, a tight CAM tolerance to maintain form accuracy, and a corner strategy to manage radial engagement around the pocket walls.

Using a conventional approach, the programmer programmes a 50% feed reduction in corners and runs the operation with a standard finishing tolerance. The surface finish in the straight sections is acceptable. In the corners, there is evidence of chatter and a visible surface defect that falls outside the inspection limit.

With DigitalCNC, the same NC program is analysed against the actual kinematic performance of the machine tool. The feed per tooth map shows that in the corners, the combination of machine deceleration and the tight CAM tolerance setting is producing chip thickness values below the edge rounding threshold of the tool. The 50% feed reduction has not been sufficient because the machine was already decelerating before the programmed feed change took effect.

The solution in this case involved adjusting the CAM tolerance distribution and the corner entry strategy, rather than simply reducing feed further. The re-analysed programme showed consistent chip thickness through the corner transitions. On the machine the cornering challenges were resolved without any increase in cycle time.

KEY TAKEAWAYS

• The tolerance settings in your CAM system and on your controller interact in ways that directly affect chip thickness and surface quality, particularly in corners and deep pockets. Understanding this interaction is not optional on high-value parts. 
• A 50% corner feed reduction is a workaround, not a solution. It does not account for machine deceleration, controller tolerance behaviour, or the specific geometry of the feature being cut. 
• By modelling the actual kinematic behaviour of your machine tool against your NC programme, it is possible to identify where chip thickness will fall below safe thresholds before the programme runs, and to adjust tolerance and strategy settings accordingly. 
• The result is a finishing process that is genuinely understood rather than conservatively de-risked, with predictable surface quality, better tool life, and a reliable basis for process transfer and quotation. 

 If you have ever adjusted a corner feed, tightened a tolerance, or run a trial cut to chase a surface finish problem, the chances are you were solving the right problem with incomplete information. DigitalCNC gives you the information that was missing. See what your machine is actually doing with your toolpath before it touches the part.

The post Your CAM Program Is Perfect. So Why Is the Part Not? appeared first on DigitalCNC.

The Surface Finish Problem Starts Before the First Cut

We published a piece this month that engineers keep getting wrong. Not because they lack skill, but because CAM gives them no visibility into machine behaviour.

CAM tolerance is treated as a geometry decision. It is actually a machine-specific decision, made without any data about the machine that will run the code. Too loose and you get geometric faceting. Too tight on a high-performance 5-axis machine and you saturate the controller, creating witness marks from feedrate fluctuations, not from bad geometry. The instinctive fix: adding more points only to make it worse.

In a recent aerospace case, CATIA predicted a 60-second cycle. Actual machine time: three and a half minutes. Prove-outs are expensive. Predicting feedrate behaviour before cutting is not.

Find the full article here. 

Carbide Is the New Platinum. Start Programming Like It.

If you have end mills in your tool crib right now, they are about to cost significantly more to replace. This is not a supply chain blip. It is a structural shift.

China controls roughly 82% of global tungsten production. Tightened quotas, export controls, and surging defence demand pushed prices up more than 200% in 2025 alone. One of the world’s largest carbide manufacturers announced 22% average price increases in Q2 2025. Analysts expect elevated prices to hold into 2027.

Dynamic milling, adaptive clearing, and proper entry and exit moves are no longer just best practice. They are cost control. Every prove-out iteration on a constrained 5-axis machine now carries a sharply higher price tag. Nothing quite focuses the mind on toolpath quality like watching a £40 end mill snap on an incorrect entry move.

The full article is now available here. Read it before your next tool order.

We were in Belfast!

When the market moves this fast, strategy cannot live only at the top. So we took the whole DigitalCNC team to Belfast for a full day reset, hosted by Invest Northern Ireland.

It was a day of honest conversations, clear thinking, and getting everyone around the same table with the same picture of where we are going. That kind of alignment does not happen on a Zoom call.

We came back sharper, more connected as a team and excited about what is ahead. Here are some snapshots from the trip.

Our Webinars are Now Available on Demand

Our webinar recordings are now available on the website, featuring AMRC experts alongside our CEO, Dr Rob Ward.

Conquering the Digital Gap: Advanced CAM, virtual machining, and where DigitalCNC fits in a precision machining workflow. Tune in now.

High-Speed, High-Value: Adaptive milling strategies and high-speed machining validation. Watch it here.

What Is Your CAM System Not Telling You About Your 5-Axis Toolpath?

Simultaneous 5-axis motion produces feedrate fluctuations that leave witness marks and surface defects. CAM cannot show you where or when they will occur. You will not see them until the part is off the machine and the damage is already done.

On 9th April, Rob will be joined by Dr Robert Carroll and Bethany Cousins, from the AMRC, to tackle this problem head on. Where slowdown zones appear, why 3+2 is often the faster and cleaner choice, and how to validate offline before the toolpath goes anywhere near a spindle.

If you programme 5-axis toolpaths or investigate surface defects, this session was built for the problems you deal with every day. Here’s the link to register, if you haven’t already.

Our CTO Had 30 Seconds to Answer One Question. Watch What He Does With Them.

How does DigitalCNC support turnkey solutions? Our CTO, David Wilkinson answered it without a slide deck, a preamble, or a disclaimer. Short, direct, and worth a few seconds of your time.

Watch it here.

We Are Funding the Research That Closes the Gap Between Simulation and Reality

We are proud to be sponsoring an Engineering Doctorate at the University of Sheffield. The mission: build physics-informed machine learning models using Bayesian inference and Gaussian processes that give CAM software a genuine understanding of how individual machines behave.

This is not a thesis that gets shelved. The researcher works embedded with DigitalCNC and the AMRC throughout, with direct involvement from Rob and David. The work feeds into the product.

As Rob put it: “We are unashamedly nerds. We get excited about Gaussian processes and Bayesian inference. If your idea of a good conversation involves machining physics and at least one tangent into science fiction, you will fit right in.”

£28k tax-free stipend. £35k research budget. CEng support. Apply by 22nd April. Open to UK Home students. Share this if you know the right person.

Yorkshire 42 Under 42. Rob’s Response: Right, Back to Work.

Rob was named in Yorkshire’s 42 Under 42 for 2026. His ambition is unchanged: get more world-class research out of universities and into industry. Not trend-chasing, not rebranding existing tools with AI language. Solving hard problems for manufacturers who cannot afford to get it wrong.

DigitalCNC exists because decades of machining research at the University of Sheffield and the AMRC deserved to become a product. This is a reminder of why that work matters.

Congratulations, Rob!

Read more about the announcement here.

April Is Loaded. Stay Close.

Big partnerships. Bigger announcements. The team is genuinely excited about what is coming and we are not in the habit of overpromising.

Follow our LinkedIn page. The next chapter is already in motion.

The post March in Motion: Inside DigitalCNC’s Monthly Highlights appeared first on DigitalCNC.

Carbide pricing has been climbing steadily, but what many shops are only just beginning to feel is the sharp upward pressure that has been building over the last several months. This is not a temporary blip. It is a structural shift in the market, and if you are not already factoring it into your pricing and programming strategy, it is time to start.

What Is Driving the Price Surge?

Several forces have converged at once to tighten supply and push costs upward. China accounts for roughly 82% of global tungsten production, and since 2023 Beijing has been tightening mining output controls, with 2025 quotas cut by a further 6.45%. The knock-on effect has been severe: according to Fastmarkets, Chinese tungsten product prices surged by more than 200% across 2025 alone, driven by export controls, national security concerns, and surging demand from the semiconductor and defence industries.

Military demand is a significant and growing factor. Global military procurement of tungsten materials rose from 2,200 tons in 2024 to an estimated 3,000 tons in 2025, fuelled by demand for armour-piercing munitions and high-performance defence components. Add to this a broad uptick in manufacturing activity and sustained oil and gas production requiring precision-machined components, and you have a market under serious strain.

The impact on tooling brands is already visible. One of the world’s largest carbide manufacturers, announced average price increases of 22% on tool products in Q2 2025. For smaller manufacturers the situation is even more acute, with around 30% of small and medium-sized tooling enterprises in Europe reported to have suspended taking new orders entirely due to supply shortages and cost pressures.

The Margin Problem Nobody Is Talking About Yet

This is still somewhat under the radar for many end customers and even some shop owners. But that window is closing fast. Analysts expect prices to remain elevated well into 2027, with current levels now considered the new market floor. Carbide tooling costs are going to start quietly eroding your margins if you are not proactive about it.

The question every shop owner needs to be asking right now is twofold. First, can these costs be passed along to your customers? Second, are your customers even aware that this is happening? If the answer to either question is no, then education becomes part of your job. Customers who understand the supply chain dynamics are far more likely to accept necessary price adjustments than those who receive an unexplained invoice increase.

Toolpath Optimisation as a Cost Control Strategy

This is precisely why the content of our Webinar 3 is so timely. Efficient 5-axis toolpath programming is no longer just about cycle time or surface finish. It is a direct lever on your tooling budget.

Optimised toolpaths reduce unnecessary cutting forces, manage heat more effectively, and avoid the kinds of sudden directional changes that cause premature tool wear. When you extend the usable life of each tool, you are buying fewer replacements. When your programming is disciplined and organised, you avoid over-ordering and reduce the waste that comes from damaged or lost tooling buried in a disorganised crib.

Three strategies in particular are worth building into your programming workflow right now.

Dynamic milling keeps the tool moving in continuous, flowing arcs rather than hard corners, maintaining a consistent chip load throughout the cut. This dramatically reduces the heat spikes that degrade carbide edges, and modern CAM packages are getting better and making it easier to implement across 5-axis toolpaths.

Adaptive clearing takes this a step further by continuously adjusting the radial engagement angle as the tool moves through the material. Rather than allowing the cutter to suddenly bite deeper into a corner or step, the toolpath compensates in real time. The result is a far more even load on the cutting edge across the entire operation, which translates directly into longer tool life.

Smooth entry and exit strategies are often overlooked but are equally important. Plunging straight into material is one of the fastest ways to chip or fracture a carbide tool. Using helical or ramped entry moves distributes the initial cutting load gradually, protecting the tip and edge from the shock that causes premature failure.

Investing time in these programming practices is one of the most practical responses a shop can make to a volatile materials market. The cost of carbide may be outside your control. How long each tool lasts is not.

Now is the time to treat every cutting edge as the valuable asset it is increasingly becoming.

Simulate Before You Cut

One of the smartest ways to build confidence in these techniques before committing to real tooling and material is to use DigitalCNC to model how your machine will actually behave. Rather than learning through trial and error on the shop floor, where mistakes cost you the very carbide you are trying to protect, DigitalCNC lets you see exactly how dynamic milling paths, adaptive clearing strategies, and entry and exit moves will play out on your specific machine. You can identify potential issues, refine your approach, and step up to the machine knowing your toolpath has already been validated. In a climate where every tool has a sharply rising price tag attached to it, that kind of confidence is not just reassuring. It is good business.

The post The Hidden Cost Crisis Coming: Carbide Pricing and Why Toolpath Optimisation Has Never Mattered More appeared first on DigitalCNC.

What the Tolerance Setting Actually Does

Every CAM system converts smooth CAD geometry into point-to-point linear segments. The tolerance setting controls how closely those segments approximate the original curve. Too loose, and the segments are large enough to leave visible geometric faceting on the surface. Too tight, and a different problem emerges.

On a modern high-performance 5-axis machine, the CNC reproduces every one of those linear segments exactly as programmed. Each segment ends in a direction change, forcing the axes to decelerate, change direction, and accelerate again. On the complex multi-axis moves that define 5-axis finishing passes, this happens continuously across the entire surface. Beyond a certain density of points, rather than improving surface quality, the toolpath creates witness lines, not from geometric error, but from the machine’s dynamic response to a continuous stream of micro-deceleration events.

These are two distinct problems with opposite remedies. Geometric faceting is solved by tightening tolerance. Controller throughput saturation is made worse by it. Setting tolerance correctly means finding the right balance for a specific part geometry on a specific machine.

Why This Is Specifically a 5-Axis Problem

In 5-axis simultaneous machining, the rotary axes sweep through arcs while the linear axes move at the same time. The tool tip trajectory is not simply the sum of those movements. The CNC has no information about the desired surface profile, only the axis positions it has been given. It cannot compensate for kinematic errors introduced as the rotary axes move between programmed points.

This means a toolpath that looks perfect in CAM simulation can still produce a surface that reflects how the machine moved between points, rather than the geometry the designer intended. Siemens’ guidance makes this explicit: chord tolerance must be set in the context of the machine’s dynamics. The tolerance value in CAM is not just a geometry decision. It is a machine-specific decision, being made without any data about the specific machine that will cut the part.

The Instinctive Fix Makes It Worse

When engineers see surface striations caused by controller throughput issues, the natural response is to tighten the tolerance further. More points, smaller segments. In this specific scenario, that is the wrong direction. Adding more points multiplies the direction changes. The constant acceleration and deceleration cycles now occur at higher frequency, programme files grow significantly, and the CNC spends more processing resource handling small blocks.

It is worth noting that all major CNC manufacturers have developed controller-side features to address this problem. FANUC’s Nano Smoothing, Siemens SINUMERIK’s COMPCAD, and Heidenhain’s Advanced Dynamic Prediction can all smooth dense linear toolpaths intelligently without sacrificing accuracy. These are standard features on modern high-end machines and should be understood and configured correctly as part of any 5-axis finishing strategy. However, enabling these features is not a substitute for selecting an appropriate tolerance in the first place. If the tolerance is poorly matched to the machine’s dynamics, you are relying on the controller to compensate for a decision that could have been made correctly upstream.

The Business Cost

Prove-outs on constrained 5-axis capacity run at several hundreds of pounds per hour. When surface quality issues emerge, the response is to iterate the toolpath, regenerate NC code, re-post, and run again. Each iteration consumes machine time that produces no saleable parts. Testing five different strategies on a one-hour process can cost £1,500 in machine time alone.

The deeper problem is that there is currently no reliable way to predict whether a tolerance setting will produce an acceptable surface before the machine runs. Prove-outs become the validation mechanism, and prove-outs are expensive.

DigitalCNC Solves This Before You Cut

The tolerance setting problem is a visibility problem. CAM engineers make decisions about how a toolpath will behave on a machine they cannot see from within their CAM environment.

DigitalCNC closes that gap. Working as a plugin inside CAM, it uses machine-specific kinematic data to predict actual feedrate behaviour across the surface in under a second. Engineers can see exactly where the controller will decelerate through a critical surface region, adjust the strategy in CAM, and re-analyse immediately, without touching the machine.

In a recent aerospace case study, CATIA predicted a 60-second cycle. The actual machine time was three and a half minutes. DigitalCNC predicted the real feedrate behaviour before cutting, the CAM strategy was adjusted, eliminating the need for multiple prove-outs and reducing engineering decision time from days to hours.

The right CAM tolerance is a part and machine-specific decision. DigitalCNC gives you the data to make it correctly, first time.

Join our webinar: What Your CAM Software Doesn’t Tell You About 5-Axis Machining. Wednesday 18th March, 15:00 GMT. Register here

The post Why Getting CAM Tolerance Wrong Ruins Your 5-Axis Surface Finish appeared first on DigitalCNC.

February was a strong delivery month for DigitalCNC, with product improvements focused on customer success, accuracy and clearer communication. The updates in this release are designed to reduce friction in day-to-day use while strengthening the feedback loop between customers and our product teams.

Faster Issue Resolution

A key addition this month is in application bug reporting. Users can now report issues directly from within DigitalCNC, without leaving their workflow. This enables faster reporting with better technical context, allowing our support and engineering teams to diagnose and resolve issues more efficiently. For customers, this means fewer interruptions and greater confidence in day to day operations.

Customer Led Roadmap

We have also introduced in application feature suggestions, making it easier for customers to share ideas and feedback as they work. This ensures that user insight feeds directly into our roadmap and prioritisation decisions. By capturing feedback at the point of use, we are able to focus development effort on changes that deliver the greatest operational value.

Seamless CAM Integration

In addition, this release adds compatibility with Designcenter NX 2512, allowing DigitalCNC to be run directly from Designcenter. This improves integration with existing CAM environments and supports earlier validation of real machine behaviour, helping teams make informed decisions before cutting metal.

Together, these updates reinforce our commitment to building a product that combines technical accuracy with practical usability, shaped by the realities of modern CNC operations.

From the Founder’s Desk

Where CAM Meets Machine Reality

In his recent article, our founder, Dr. Rob Ward, examines why CAM predictions often fail to reflect real machine performance, using practical CNC examples to challenge common assumptions around machining strategy, cycle time and quality outcomes.

The articles explore:

• Why continuous 5 axis machining can run slower than 3 plus 2 on real machines

• How kinematic limits and controller behaviour impact feedrates and cycle times

• The hidden quality risks caused by feedrate variation and machine dynamics

• When simultaneous 5 axis machining is the right choice and when it is not

• The real cost of trial and error when machining decisions are based on assumptions

Read more: Why 3 plus 2 Is Often Faster Than 5 Axis

Where Decisions Should Happen

In this article, our founder examines why many of the most costly issues in aerospace machining are discovered too late in the process. Drawing on systems engineering principles and real manufacturing workflows, he explains how pushing critical insight upstream fundamentally changes the economics of machining decisions and reduces risk before production begins.

The article covers:

• Why the cost of change increases sharply once machining reaches the shop floor

• How gaps in information timing, not engineering capability, drive late stage failures

• Why CAM programmers are forced to make high impact decisions without machine specific data

• How upstream access to machine behaviour shifts decisions to when they are still affordable

• Why early insight improves productivity, quality and risk management in aerospace programmes

Read more: What Can We Push Left? The Case for Upstream Data in Aerospace Machining

From Our Webinar Series

This month, we also hosted a webinar on aerospace machining, focused on dynamic milling strategies and real-world decision-making. The session was led by our CEO, Dr. Rob Ward and joined by Ryan Fletcher and Joe Berner from the Advanced Manufacturing Research Centre (AMRC), who shared practical insights from aerospace machining programmes.

The session explored:

• Key challenges in designing toolpaths for aluminium and titanium aerostructures

• The impact of machine kinematics, feedrates and toolpath selection on performance

• A real aerospace case study showing how strategy choice affects cycle time and consistency

• Why early access to accurate machine data improves decision making and reduces risk

• How digital tools can be integrated into existing CAM workflows with minimal disruption

Watch the recording: High-Speed, High-Value: Mastering Adaptive Milling Strategies with DigitalCNC– Shared screen with speaker view

Passcode: !sxv6@mY

Beyond the office!

At a recent industry event hosted by Seco Tools, our CEO shared a clear view on the role of AI in manufacturing. The real opportunity is not AI in isolation, but AI grounded in a deep understanding of machining processes and validated against real data. Knowing what actually happens when metal is cut is what enables engineers to make better decisions earlier and with confidence. With organisations such as the Advanced Manufacturing Research Centre (AMRC) building this foundation, the challenge now is adoption and alignment across industry, investors and policy to turn world-class research into real commercial outcomes.

Photo courtesy: Seco Tools

Read more here.

The post Inside DigitalCNC: February 2026 Updates appeared first on DigitalCNC.

What can we push left?

The specifics change. The principle never does.

The Cost of Change Curve

Systems engineers have known for decades that the cost of changing something increases exponentially the further downstream you make that change. It is one of the most well-evidenced principles in engineering management, and it applies whether you are designing an aircraft, developing software, or managing a complex machining programme.

The logic is straightforward. Early in any process, decisions are provisional. The commitments are limited and the options are open. Change something at the design stage and the cost might be a few hours of an engineer’s time. Change the same thing on the shop floor and you are looking at scrapped parts, unplanned downtime, rescheduled operations, and the downstream ripple effects that rarely show up cleanly on a cost report but are felt acutely by everyone involved.

In aerospace manufacturing, where part complexity is high, tolerances are tight, and material costs are significant, this dynamic is particularly unforgiving. And yet the dominant model in the industry still pushes the moment of discovery dangerously far to the right.

How Problems Reach the Shop Floor

Consider the typical workflow for a complex aerospace component. A CAM programmer develops toolpath strategies, drawing on experience, established post-processors, and verification software. The programme is verified to a reasonable degree of confidence and released to the shop floor.

Then machining begins.

And it is here (often hours or days into a process that may have taken weeks to plan) that the real picture emerges. A strategy that once looked sound in theory doesn’t perform under actual cutting conditions. Feeds and speeds that appeared achievable cannot deliver the minimum chip thickness. Tooling that was specified at the programming stage turns out to be a poor choice for the actual cutting conditions encountered.

None of this is a failure of competence. It is a failure of information timing. The programmer made the best decisions they could with the data available at the time. The problem is that critical data (about how that machine, that toolpath and those parameters will perform) was not available at the point when decisions were still cheap to change.

By the time it is available, it is not cheap at all.

Upstream Data Changes the Economics

The concept of shifting decisions to the left is not new. Lean manufacturing, concurrent engineering, and digital twin methodologies have all been working at this problem for years. What has changed is the availability of the data required to make early decisions with genuine confidence rather than informed intuition.

This is precisely the challenge that DigitalCNC was built to address.

By giving CAM programmers access to machine specific data at the point of programming (before a single cut is made), the economics of the cost curve shifts fundamentally. The questions that would previously have been answered only by the machine now have answers at the desk. Will this strategy achieve the required chip thickness? Where are the margins, and how sensitive is the outcome to changes in feedrate? These are not small questions. In large-part aerostructure machining, where a single component might represent significant material value and a machining cycle of many hours, getting the answers wrong at the shop floor stage is costly in ways that compound quickly.

Getting the answers right at the programming stage costs very little by comparison.

What This Means in Practice

The immediate effect of upstream data is faster, more confident programme development. CAM programmers are not replacing their expertise. They are augmenting it with a layer of process intelligence that previously only existed in the collective memory of experienced machinists and in post-machining analysis of completed runs.

But the second-order effects are where the real value accumulates.

When problems are surfaced upstream, they inform not just the current programme but the broader knowledge base of what works and what does not for a given strategy, tooling family, or machine configuration. Over time, this builds a data asset that compounds: each programme making the next one faster and more reliable.

There is also a risk reduction dimension that matters particularly in the aerospace supply chain. Late-stage process failures do not just cost money. They affect delivery schedules, they consume engineering resources that should be focused elsewhere, and in the most serious cases they create quality exposure that has consequences well beyond the immediate programme. Upstream data is, in this sense, a risk management tool as much as a productivity one.

The Broader Principle

Pushing left is ultimately about the relationship between information and decision-making. Every engineering process involves a sequence of decisions, and the quality of those decisions is bounded by the quality of the information available when they are made.

The goal is not to eliminate uncertainty (that is not achievable in complex manufacturing environments). The goal is to ensure that the right information is available at the right point in the process, so that the decisions that are most consequential are made when they are still affordable to revisit.

In aerospace machining, that point is at the CAM programmer’s desk. Not the shop floor.

Push Left Today

If you are responsible for machining programme development in the aerospace sector, it is worth asking the same question that has shaped my thinking across every engineering environment I have worked in. Where in your process are problems being discovered that could have been surfaced earlier? What decisions are being made on the shop floor that could be made at the desk, with better data, at a fraction of the cost?

What can you push left today?

Dr Rob Ward, CEO, DigitalCNC

DigitalCNC provides CAM programmers with machining intelligence at the point of programming, enabling better decisions upstream and fewer problems downstream. If you would like to understand how this applies to your specific machining challenges, get in touch.

The post What Can We Push Left? The Case for Upstream Data in Aerospace Machining appeared first on DigitalCNC.

This isn’t a one-off. It’s the gap between CAM predictions and machine reality.

The Hidden Slowdown

Continuous 5-axis demands constant coordination of all five axes. When programmed feedrates hit kinematic limits during tool reorientation, the controller decelerates to maintain accuracy. CAM doesn’t see these slowdowns.

Siemens knows that you get “frequently longer machining times due to compensatory movements of the kinematics” and FANUC says that “because of the constant acceleration and deceleration of the axis and the fact that the abrupt changes cause machine shock and vibration, feedrates are limited.”

However, when machining in 3+2, the rotary axes are locked. Only three linear axes move during material removal. No rotary kinematics forcing deceleration. More consistent feedrates throughout.

Quality Matters More

Beyond cycle time, there’s a bigger problem. Feedrate fluctuations in 5-axis leave witness marks and surface artefacts.

On thin-walled aerospace parts, any feedrate variation shows as a finish defect. For medical devices, it impacts final quality. For precision components, it means rejection.

Siemens is explicit: “Even the smallest jumps in deceleration and acceleration can result in surface defects (e.g. chatter marks).” FANUC describes it exactly: “The modern CNC will faithfully reproduce those small segments generating witness lines on the part.”

With 3+2, you eliminate feedrate variations from rotary axis dynamics. Locked tool axis means more consistent chip load, improved surface finish and longer tool life.

When 5-Axis Works

This isn’t anti-5-axis. It’s about choosing correctly. True simultaneous 5-axis excels for sculptured surfaces requiring continuous tool tilt, deep cavities needing collision avoidance during cutting, and parts demanding continuous rotation for accessibility.

But for many components (especially those decomposable into distinct orientations), 3+2 delivers better results. Both Siemens and FANUC recommend: “As much as possible 3-, 3+1- and 3+2-axis roughing/semi-finishing. 5-axis simultaneous milling only for the finishing.”

The Real Cost of Guessing

CAM can’t tell you which approach performs better on your specific machine. Without understanding machine kinematic capabilities and toolpath interaction, you’re programming blind.

On a recent aerospace use case: CAM predicted 40 seconds for a deep-pocket aluminium operation. Actual time? Three minutes. That’s 4.5x error. Not programmer mistake. CAM simply can’t account for machine kinematic limits.

FANUC testing shows the scale of the problem. Their comparison data: setup dropped from 2.5 hours to 30 minutes, cycle time reduced from 80 minutes to 40 minutes, surface finish improved from rough to smooth. Same machine, same part, different approach.

Testing strategies costs real money. Each test consumes machine time, material, programming effort. For aerospace manufacturers, testing five strategies on a one-hour process can cost £1,500 in machine time while producing nothing.

Make Better Decisions

The path forward isn’t avoiding 5-axis. It’s knowing when to use it. Recognising the gap between CAM predictions and machine performance lets you make programming decisions that optimise both cycle time and quality.

The question isn’t whether your machine has 5-axis capability. It’s whether continuous 5-axis is the right strategy for this part, on this machine, with these quality requirements.

Want to learn more about bridging the gap between CAM and machine performance? Join our upcoming webinar and learn more about What Your CAM Software Doesn’t Tell You About 5-Axis Machining.

The post Why 3+2 Is Often Faster Than 5-Axis appeared first on DigitalCNC.