Frequently Asked Questions

Evolutionary algorithms are software programs that simulate evolution to solve a problem. Typically a number of starting designs are generated randomly following parameters, or directly provided by the user. Then the algorithm will check each design to a set of criteria to calculate their fitness. Only the fittest designs will go to he next round, where the designs are altered slightly (mutation) or combined with other designs. Then the process continues iteratively for many generations. This will produce new emerging designs that will gradually become fitter than the starting designs.

By selecting different starting designs and a different way to measure fitness, the result can vary.

Additive manufacturing is the same as 3D printing.

A white label product is a generic product manufactured by one company that can be rebranded and sold as if it were made by another company. For example Korall offers our products as white label, meaning they can be deployed under customer domain, with customer logo and branding (digital-inventory.yourcompany.com).

Design optimization refers to redesigning existing parts to improve their characteristics.

This may include part consolidation and removing excess material through for example topology optimization . This is particularly useful when using 3D printing .

For products where mass is critical, reducing weight helps cut down fuel consumption and CO2 emissions. For high-performance applications, we identify design improvements to maximize efficiency.

3D scanning is the process of capturing a physical object's shape and appearance to create a digital 3D model. It works by collecting data points on an object's surface, which are then used to build a three-dimensional digital replica. This model can be used for a wide range of applications, such as 3D printing, reverse engineering, design, and virtual reality.

In traditional design methodology, we would collect requirements from the project and make a design to meet those requirements. We would then create the necessary tooling to possibly mass manufacture this part, and then store the end result in a warehouse.

Whenever the requirements change, even by a little, the process would have to repeat. With the advent of additive manufacturing, digital inventories and more advanced design tools, we can now create one single design that has built in parameters that can be changed after the design is complete. For example, a gear can have parameters for the bore diameter, number of teeth and depth.

One parametric gear design can thus result in thousands of resulting designs for parts, often referred to as a family of parts.

Part consolidation is the process of combining many parts into one in an effort to remove excess material, assembly and unnecessary complexity. In traditional manufacturing, there are many limitations to which designs are feasible to manufacture. With the advent of 3D printing, many of these limitations are no longer present, and so it makes sense to re-design parts. Where once an assembly of multiple parts was the only alternative, now we could combine the designs into one and get rid of a lot of wasteful complexity and material. You can get a part that is cheaper to manufacture, weighs less, is structurally stronger and generally better.

With traditional manufacturing, the time and cost of designing and manufacturing a new part was substantial, and so we would produce a large number of parts and keep them in a warehouse.

In the world of additive manufacturing, the time and cost of designing and producing a new part is nearly insignificant in comparison while the potential number of designs possible is endless. This makes it infeasible to keep a stock of parts, but rather we want to keep a stock of designs in a database. This is the idea behind a digital inventory. Instead of finding a physical part in a warehouse, we find the digital counterpart in a database and print it on demand.

Topology optimization is an algorithmic method of design optimization where a part design is iteratively optimized by detecting and removing material that is not needed structurally by the model, leaving a finished design that is very minimalistic in terms of weight and material, while still being strong in the right ways. This lends itself particularly well to 3d printing, since the resulting design might be difficult to manufacture using traditional methods.

FDM stands for Fused Deposition Modeling and it is the most common method of 3D printing. It works by moving a nozzle around while depositing material onto an object

SLS stands for Selective Laser Sintering and is a method for 3D printing that fires a laser into a container full of compacted powder. Where the laser hits, the powder will melt (sinter) and form into a solid layer of the desired model. For each step, new powder will be added on top, and the laser will fuse the powder of each layer into a single object in 3D.

SLS can be used with both metal and polymer materials, and has many benefits when compared to other 3D printing methods such as FDM.

Generative design is the collection of design practices related to creating designs using algorithms rather than explicitly defined features. For example, you could use an algorithm that simulates the way a tree grows within parameters. By changing the parameters you can run the algorithm iteratively to see different emerging designs and select the one that matches your requirements best.

Generative design is often combined with machine learning and evolutionary algorithms to further improve the capabilities of the process. Machine learning (a form of AI) can be used to incorporate advanced decisions into the algorithm, or statistical models approximating the real world. Evolutionary algorithms can be used to adapt to constraints set by the designer. For example, to find a structure can hold the load without wasting material.

Design automation are the collection of design practices that allows for automating the design process. Typically this is done through the use of parametric models. For example, you can create a cad model of a part with built-in parameters, such as critical dimensions. Then whenever we need a new part, instead of starting from scratch, now a non-skilled worker can select the appropriate values for the parameters according to the requirements to generate a single combination part automatically.

All the parts that derive from a single parametric model are often referred to as a family of parts.

A family of parts in the context of parametric design refers to a single parametric model being used to produce the blueprint of many distinct parts that all vary only with the values of the parameters inherent to that model.

For example, if you have a parametric model of a gear, where the parameters correspond to critical dimensions such as

• Number of teeth

• Tooth height

• Bore diameter

• Gear outer diameter

Then no matter the combination of parameters you chose, the resulting part will belong to the "gear" family and the part will be defined uniquely in the family by the parameters themselves.

3D printing, also known as additive manufacturing, is a process that allows you to create an object from your design in a computer.

There are many methods to achieve this, but the most basic builds the object by depositing material layer by layer. By adding material, such as plastic, metal, or resins, in thin layers stacked on top of each other, we can gradually form an object from the ground up. This is different from traditional "subtractive" manufacturing, which often involves cutting or removing material from a a n original piece of stock material.