Guide to using SolidCAM to generate 2D toolpaths (pocket+profile) with CNC router

This first video shows the following steps:

  1. creating a basic part by importing a .dxf file and extruding it 18mm deep
  2. creating a co-ordinate system in Solidworks that can be used by SolidCAM
  3. defining the settings for the SolidCAM operations (co-ordinate system, part, stock)
  4. defining the pocket steps (and copying them with updated geometry selection)
  5. defining the profile step
  6. simulating the toolpath, with checks (stepdown amount, cut inside/outside, order of operations)
  7. exporting the g-code using the ‘router’ post-processor (.ncp file for this machine).

and this second video shows how to load the .ncp file with the g-code into the ‘Remote’ software which runs the machine. We also train the software to set XYZ0.

Guide to using Cut2D to generate toolpath from .dxf file to cut with CNC router

This first video shows how we’ve used the Cut2D software to load a .dxf file and generate the g-code using the following steps:

  1. Job setup including the location of the origin and material size
  2. Importing the vector files from file (.dxf), then joining the open vectors and move/scale/rotate
  3. Setup the material and rapid gaps
  4. Toolpath operations (2D profile toolpath, pocket, drill, add tabs)
  5. Edit and calculate the toolpaths
  6. Preview and save the toolpaths

and this second video shows how to load the .ncp file with the g-code into the ‘Remote’ software which runs the machine. We also train the software to set XYZ0.

Guide to using TensorFlow machine learning to identifying classifications from images

I’ve just come across TensorFlow, an open source machine learning platform.

Here’s one of their first tutorials. It focuses on basic classification of images, identifying the difference between various types of fashion items (shoes, handbags, t-shirts etc).

I wonder if this can be included into mods, so that the traces and outline can be cut from a single programme (with automatic toolchange).


Guide to setting up the PID Eurotherm controller to create composite curing cycles

This video shows how to set up the PID controller to carry out a curing cycle using the eurotherm controller. In this particular example we are looking at the standard 120°C curing cycle over 9 hours. Firstly we set the ramping time units (in most cases we use minutes), then we set the dwell time units (in most cases we use hours) . Each segment of the program starts with setting the time to ramp, then the set temperature. Then the dwelltime is defined. Then segment two will start.

Reverse engineering products

The exercise of reverse engineering products involves taking things apart. Shown here is a mounted disassembled bike (from Todd McLellan’s “Things come apart” image exhibition and now a book – see references). It’s amazing when you see the object in this form – there are so many parts, so many materials, so much design! Sometimes this can be done elegantly with a screwdriver, othertimes it needs prizing with a screwdriver, othertimes it’s an act of destruction by cutting things in half (or quarter) with a blade or saw, othertimes things can be unpicked, unlocked, unsnapped, or even thrown to the ground to be smashed (although only as a last resort!). You can take apart pretty much anything, but really good products to reverse engineer are those whose insides we’re not used to…those products we may be familiar with on the outside but whose insides are a mystery. Examples include remote controllers, cameras, shoes, phones, gearboxes, game consoles, joysticks, toys, other electrical goods like printers, irons, toasters, speakers, computers, VCRs (do they still exist?) etc. Ideally use something that’s defunkt, that doesn’t work anymore, so if you break it (or can’t reassemble it!!) then you’re not too bothered. Freecycle is great, Ebay can be good too, so can charity shops, skips, or ask family and friends for broken things.

When reverse-engineering a product (and I recommend doing this often! – it’s a great way to learn), you should consider two things:

Firstly, consider what your objectives are. There’s lots to learn by taking something apart, but what are you specifically after? Try to refer this to Design for Manufacture and Assembly (DFMA) principles by Boothroyd et al (2010) where possible. Below are a list of things SOME that you might want to consider.

Secondly, consider what you’ll need to disassemble the product. Will you need specialist tools? will you require personal protective equipment (PPE) in the form of gloves, goggles, labcoat? will you need a cutting mat? will it be messy? will you need lots of space? how will you lay out the object to the best effect? will you need to re-assemble the object, in which case how will you layout and/or label the objects? when you record the process, how will you do this? pen/paper? photographs? if so, how will you stage this to get the best effect using a tripod? or how will you keep your camera/phone still? and with good lighting? with a suitable clean backdrop? will you be able to capture it all in one go or will you need to piece images together in photoshop? (for the big products – like the tractor – Todd McLellan (2013) did just this).

A sample list of things to take to a reverse engineering party:

  • pen and paper
  • a decent camera (or camera phone) and something to mount it on (ideally tripod) – also consider how you’ll take pics to maximise the visual effect (a nice matt white background is always good!)
  • white sheets of paper taped together to lay things out on
  • steel and plastic rulers
  • screwdrivers (normal size and small terminal size), both phillips and flat head
  • other flat tools for prizing open troublesome products
  • allen keys (and other specialist tools for unfastening?)
  • sharp stanley knives and cutting mats
  • blue tak or superglue (to hold small parts on the table)
  • sticky tape and masking tape
  • digital callipers
  • scissors
  • possibly board (painted white) and pins/screws/wire/fishingline to mount things permanently!

For the super keen, mount all the parts neatly onto a piece of painted board (possibly with a small label) and we’ll put it up in the studio!

Nikko radio controlled ford fiesta rs_Chloe_fong

General assembly/disassembly considerations (related to Design for Assembly – DFA guidelines):

  • how many parts does the product have? often products have individual part numbers actually shaped into an internal surface.
  • how many of the parts are standard? how many are “designed”?
  • how many different materials are actually used?
  • what is the assembly process?
  • to what extent is the product able to be disassembled at end of life?
  • what type of fit is there for the product? is it possible to measure the tolerances for any mating components to understand how much clearance or overlap is there?

Energy,heat and power considerations:

  • how much power does the product draw? where does that power get used? by what components and how much for each?
  • does it use a battery? what type? how much energy does it store? how big? how heavy?
  • identify heating elements, heat sinks, fans, and consider how the heat is managed within or around the part
  • are there any motors? if so what type, how fast? how much torque? and power?
  • what about gears? what type are they? what are the gear ratios? how many teeth do they have?
  • are there other pulleys or drives used? if so how are they used?
  • are there any springs? what type? how are they used?
  • other mechanical?

Plastic specific considerations:

  • how was the part made? is there any evidence on the part to give you a hint? this could be a parting line (showing where the mould cavities meet), or an ejector pin mark.
  • what material is it? often it’s labelled.
  • what draft angles are used? you’ll need to measure this!
  • has the product been glued, ultrasonically welded, rivetted, screwed, snap fit, other? look for clues before you take it apart.
  • what features are included internally and why? consider tabs, ribs to stiffen, bosses to screw into, slots to guide, other mounting features, strain relief etc.
  • what are the material wall thicknesses? these vary from main walls to ribs, bosses, other features. There are recommended wall thickness for different plastics. It’s always good to check that the material you’re looking at fits with these guidelines!
  • what are the thick to thin wall transitions like? are they suitable? are there voids or sink marks present on the part?
  • what are the surface finishes like on the surfaces of the part? these should give you a hint about the quality of the mould, or the requirements of the part itself. They can also help to show or hide any manufacturing problems (e.g. masking sink marks).
  • is there a mix of flat or curved surfaces? does the quality of their surface finish differ?
  • for any living hinges what’s the material thickness? and how have they designed this into the part bearing in mind moulding processes? this is clever indeed!
  • for any snap fits, what is the shape of the cantilver that “bends”? does it taper or is it straight? how does this fit with the design guidelines for snap fit? (or not!) are there signs of strain at the join?
  • if it’s a moulded part, is it a straight pull mould (ie where the mould can be pulled straight from one side to leave the part – with no undercuts), or would inserts be required to allow for undercuts to be produced? often there is evidence in the sign of parting lines to indicate various inserts for the mould.

Phillips Norelco Shaver Tristan Rose

Electronic specific considerations:

  • try to identify what electronic components are present?
  • try to identify which are sensors and which are actuators?
  • how are things mounted to a) each other, b) the case?
  • for any circuitboards, are the components surface mounted or through hole mounted?
  • are the technical specifications of the components obvious? what are they?
  • what does the main circuit look like? can you draw this? for battery driven products, are the batteries in series or parallel? and if so what’s the voltage?

Electrical specific considerations:

  • how has strain relief used for mains cables?
  • is there an earth? what colours are used for the wires? do they conform to relevant BS safety standards?
  • how is the wiring connected to terminals/pins?
  • is the casing suitably insulative?
  • is it waterproof? it’s worth looking into IP (ingress protection) requirements for electrical products.

Sheet metal considerations:

  • what processes were required to make the part? these could be cutting, bending, punching, rolling, stamping, etc…
  • in what order were these processes done? is there a way to tell?
  • what kind of fastening was used? welding, rivets, screws, glues, etc.
  • what kind of tools or punches were used to make certain features?
  • what are the bend radii for any bends?
  • are there any notable design features, e.g. reliefs in the corners?
  • are there any surface treatments used to finish, coat or debur the part?

Machining specific considerations:

  • what processes were required to make the part? drilling, milling, turning?
  • in what order were these processes done? is there a way to tell?
  • were any tertiary processes required to finish the part to give a particular surface finish or detail?
  • how was the object held while it was made?
  • are there any key datums, e.g. surfaces on the part?

Others? there are plenty more things to consider and to learn from these exercises!


Boothroyd, G., Dewhurst, P., and Knight, W. 2010, Product Design for Manufacture and Assembly, Third Edition, CRC press.

McLellan, T. 2013, Things Come Apart: A Teardown Manual for Modern Living. Thames & Hudson.

The Chris Hoy vs AA batteries challenge

The challenge

How many AA batteries would it take to have more energy than Chris Hoy in a sprint burst?

Let’s say we have a AA Duracel battery which is 1.5 V rated at 2500 mA.h (this is a good AA battery!).

To understand how much energy it has stored in it, we multiply 2500 mAh by 3600 (to convert the hours to seconds), which gives us 9,000,000 mA . seconds (or 9,000 Amp.seconds).

If we multiply this by 1.5 V, we get 13,500 Amp.Volt.seconds (or Joules), that’s 13.5kJ.

Remember from an earlier post that Sir Chris can churn through 92kJ before he is spent (in a full sprint)…

…that’s the equivalent of nearly 7 x AA batteries…hmmm, that doesn’t seem like much…but there’s a catch…

The AA batteries are only really rated at 50 mA…so there’s no way they could compete with that power output! They would get very hot (and may explode!).

If we drew 50 mA from each of the batteries, they would last 50 hours (50 mA x 50 hours = 2500 mA.h)…they would come in a distant second in the race if Chris Hoy finished in only 40 seconds!

Some notes…

clearly there are many assumptions here, the points really are that a) batteries have a limited current capacity that is worth being aware of right from the start, and b) Chris Hoy is an incredibly powerful individual!

Energy efficiency of kettles

Energy can not be produced or destroyed: it must go somewhere (1st law of thermodynamics). Often when we build a machine to do work, some of the work is useful (e.g moving the car forward), but some of it is not useful (e.g. making grumbling noises or getting hot).

When we talk about efficiency, we’re talking about the ratio of the useful work performed by a machine or process to the total energy expended.

Energy efficiency = useful work performed (J) / total energy expenditure (J)

Useful work to heat water

How efficient is my kettle in boiling 500mL of water?

Firstly, we work out the energy required to raise temperature of the water from 20 degrees C to 100 degrees C. For this we need the amount of water (assumed to be 0.5 kg since the density of water is very close to 1000 kg/ m3).

We also need the specific heat capacity of water (4200 J/(kg.K) ).  Note that we can use degrees C for the temperature difference (instead of Kelvin), since we’re only looking at the difference in temperature, and these scales are linearly offset from one another (i.e. an 80 degree C change is the same as an 80 Kelvin change).

So, our formula to calculate the amount of energy required to heat the 500 mL of water up from 20-100 degrees C, is

Energy = mass (kg) x Specific heat capacity (J/(kg.deg C)) x  change in temperature (deg C)


Energy  = m . Cp . Δ T = m . Cp . (T1-T2)

=0.5 kg x 4200 J/(kg.deg C) x (100-20) deg C

= 0.5 x 4200 x 80

= 168,000 J, or 168 kJ

So the energy required to heat the water is 168 kJ.

Efficiency of my Morphy Richards kettle

From an earlier post, I worked out that the kettle actually used 197 kJ to boil the kettle.

Remember that:

Energy efficiency = useful work performed (J) / total energy expenditure (J)

So in this case:

Energy efficiency = 168 kJ / 197 kJ

= 0.853, or 85.3 %

So, according to this quite simple calculation, this particular kettle is 85.3 % efficient in boiling my 500mL of water.

Some notes…

It took 74 seconds until I noticed that the water had just boiled, but the switch didn’t trip until 90 seconds…how efficient does this impact on the expected efficiency if I didn’t stop it early?

Also, what are the assumptions I’ve made in this? well first of all, I’ve assumed that the power consumption was constant throughout the heating time – and from observation of the power meter, it clearly was not constant. Ideally I should plot the power drawn over the heating time, then integrated this to get a truer value for the actual energy consumed.

I should note also that my measuring of the 500 mL of water and of the time taken was all rather crude, and to my knowledge the power meter is not particularly accurate!

We can calculate the energy to boil water, and the energy consumed by a kettle, but what about making toast? Again, we can work out how much energy is used, but how much energy is needed to toast bread? This is somewhat more complex…in these cases we need to take other measurements, of temperature for instance to work out what might be relevant for that particular product case. Here’s an example of a study looking into the heat transfer in hair dryers and irons, and here’s an article about a study that analysed the power/temperature conditions used to make improve toast making.


Calculating energy usage

Human power and energy

I like to start my analysis of power and energy with human power since it is something we can relate directly to. Let’s start with some definitions.

Energy is the ability to carry out work. It is measured in Joules (SI units).

Power is the rate of doing work. It is measured in Watts (SI units).

When Sir Chris Hoy was in his prime, on his track bike apparently he had a max power output of 2300W, but he can only sustain this for 40 seconds, how much energy does he consume in that time?

Well let’s assume that he has a constant power output over this 40 seconds (or that 2300 W is the average over this time).

If this is the case, then if:

Power (Watts) = Energy (Joules) / time (seconds)

then rearranging this gives:

Energy = Power x time, so

Energy = 2300 x 40 = 92,000 Joules, or 92 kJ

Here is a graph of human power output capabilities over a range of time durations. A typical adult with a reasonable fitness level can typically average between 50 and 150 watts for an hour of vigorous exercise.

Food and energy

To put that into context, a Mars bar has 1004 kJ (240 kcal or 1 MJ) stored in it…more than 10 times as much! Check out other food calories data here or the NHS calorie page here.

To note, the energy burned by someone running a marathon is about 11,000 kJ (or 11 MJ, or 2600 kcal), which is about 11 Mars bars! It’s also interesting to note that this about the same as the typical recommended calorie intake consumed by a person in one day!

here’s a useful converter to convert from kcal (food) to kJ (standard international units for everything else).

Energy, power and appliances

OK, back to appliances….

understanding the power requirements of various appliances around the house is useful. Since I’m involved with the ‘usable’ design project, I’ve compiled a list of appliances around the house that heat. They are in a random order, have a think about how much power you think they draw from the mains and rank them from 1 (drawing the most power) to 11 (drawing the least power), then scroll further down this page for the answer.

Appliance Power (Watts) Rank
Toasted sandwich maker
Electric blanket
Milk warmer
Small oil filled radiator
Fan heater
Electric hob
Electric oven TOTAL
Electric oven

Boiling water to make tea

I measure the power output of my Morphy Richards kettle (using a meter similar to this) when boiling 500 mL of water, it’s 2664 W (that’s the average, on the sticker it is rated as between 2500-3000 W), and it takes 74 seconds to boil the water.

So in this time, let’s work out how much energy it uses to boil that water:

Energy = Power x time, so

Energy = 2664 x 74 = 197,136 Joules, or approximately 197 kJ

Appliance power output

OK, so back to our appliance list, here’s the rank order with the power requirements…

Appliance Power (Watts) Rank
Electric oven TOTAL 11,000 1
Electric oven 5,000 2
Fan heater 3,000 3
Kettle 2,500-3,000 4
Electric hob 1,700 5
Small oil filled radiator 1,500 6
Toaster 1,200 7
Toasted sandwich maker 700 8
Breadmaker 600 9
Electric blanket 100 10
Milk warmer 60 11

Clearly, how much energy these things use depends on how long we run them for. An electric blanket does not draw much power (100 W) when compared with a toaster (1200 W), but we will keep the electric blanket on for many hours, while the toaster will only be toasting for a few minutes each day. Ok, the electric blanket may not be actually heating for the entire night, but you get the point.