{"id":8633,"date":"2026-07-15T08:00:17","date_gmt":"2026-07-15T06:00:17","guid":{"rendered":"https:\/\/felsaris.com\/cfd-combustion-simulation-e-fuels-engine-development\/"},"modified":"2026-08-18T09:00:18","modified_gmt":"2026-08-18T07:00:18","slug":"cfd-combustion-simulation-e-fuels-engine-development","status":"publish","type":"post","link":"https:\/\/felsaris.com\/en\/blog\/cfd-combustion-simulation-e-fuels-engine-development\/","title":{"rendered":"CFD combustion simulation for e-fuels | 1D, 3D CFD and AI"},"content":{"rendered":"

CFD combustion simulation for e-fuels: engine development with 1D, 3D CFD and AI surrogates<\/h2>\n

When calibrating a high-performance engine for e-fuels, the CO2 balance of the fuel is only one part of the assessment. The decisive processes are local phenomena inside the combustion chamber: injection, spray breakup, wall film, mixture homogeneity, ignition, heat release, temperature distribution and emissions.
\n\u200d<\/p>\n

This is where CFD combustion simulation becomes relevant. It makes visible what global metrics often cannot show: how the fuel is distributed inside the combustion chamber, which regions are over-rich or too lean, how temperature fields develop and whether model assumptions can be supported by test bench data.
\n\u200d<\/p>\n

This article explains why e-fuels create new development questions for internal combustion engines, what role 1D system simulation, 3D CFD, reactive CFD and AI surrogates play, and why robust validation remains the reality anchor for every simulation.<\/p>\n

Why e-fuels create new development questions for internal combustion engines<\/h2>\n

E-fuels are not just another fuel label. For engine development, what matters is how the specific fuel flows through the fuel system, is injected, evaporates, mixes with air, ignites and burns.
\n\u200d<\/p>\n

For liquid fuels, relevant properties include density, viscosity, vapour pressure, boiling behaviour, surface tension, heating value and heat capacity. These properties influence injection hydraulics, droplet size spectrum, spray penetration, evaporation, wall film and the local air-fuel ratio in the combustion chamber.
\n\u200d<\/p>\n

A fuel can therefore appear compatible on a data sheet and still create new development questions in a real engine. Particularly in direct-injection high-performance engines, it is not enough to consider only the global air-fuel ratio. What matters is the local mixture at the time of ignition and how stable the combustion remains across the load spectrum.<\/p>\n

Fuel data as the starting point for simulation<\/h3>\n

A robust simulation does not begin with the solver, but with the input data. For e-fuels, material properties and evaporation behaviour are especially relevant. These include density, viscosity, boiling curve, vapour pressure, heat capacity, enthalpy of vaporisation and heating value.
\n\u200d<\/p>\n

These data directly affect the modelling. A changed boiling behaviour can influence evaporation time. A different viscosity can alter injection hydraulics and jet breakup. A different enthalpy of vaporisation can cool the charge, but can also shift local mixture formation and wall-film behaviour.
\n\u200d<\/p>\n

The first development question should therefore not be: “Can the engine run on e-fuel?”
\nThe better question is: “Which fuel properties influence which target quantities in this specific engine?”<\/p>\n

Engine behaviour as the target quantity<\/h3>\n

For engine development, fuel data are only the starting point. The decisive issue is how these data affect target quantities. These include cylinder pressure, CA50, burn duration, pressure rise rate, knock tendency, NOx, CO, HC, particulates, exhaust gas temperature, wall temperatures and durability.
\n\u200d<\/p>\n

CA50 denotes the crank angle at which 50 per cent of the fuel energy has been released. This quantity is central to burn-rate analysis because it closely links ignition, heat release and efficiency.
\n\u200d<\/p>\n

With e-fuels, this creates a typical development trade-off: a fuel may have favourable properties for knock resistance while introducing new questions around evaporation, wall film or emissions. These interactions must be represented in the model chain.<\/p>\n

Injection, spray and wall film: why mixture formation is decisive<\/h2>\n

Mixture formation is one of the most important technical levers in e-fuel engines. The fuel must be present in the combustion chamber at the right time, in the right quantity and with suitable distribution. A globally correct lambda value is not sufficient.
\n\u200d<\/p>\n

For liquid fuels, several local mechanisms interact:
\n\u200d<\/p>\n

\n\n\n\n\n\n\n\n\n\n
Mechanism<\/th>\nRelevance for e-fuel engines<\/th>\n<\/tr>\n<\/thead>\n
Nozzle flow<\/td>\ninfluences mass flow rate, jet structure and repeatability<\/td>\n<\/tr>\n
Spray breakup<\/td>\ndetermines droplet size, penetration and evaporation time<\/td>\n<\/tr>\n
Evaporation<\/td>\ninfluences local temperature, mixture formation and ignitability<\/td>\n<\/tr>\n
Wall film<\/td>\ncan influence HC, particulates, oil film and incomplete combustion<\/td>\n<\/tr>\n
Spray-wall interaction<\/td>\ncouples injector position, piston bowl, wall temperature and timing<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Especially in direct-injection engines, injection timing, injection pressure, injector position, spray cone, piston bowl, tumble and wall temperature are tightly coupled. A change in fuel can therefore make the same hardware behave differently.<\/p>\n

Spray breakup and droplet size distribution<\/h3>\n

A different fuel can produce a different droplet size distribution with the same injector. Larger droplets evaporate more slowly and can penetrate deeper into the combustion chamber. Smaller droplets evaporate faster, but follow the local flow more strongly.
\n\u200d<\/p>\n

For CFD, this means that the spray model must match the engineering question. If the objective is only a coarse assessment of mixture distribution, a simpler approach may be sufficient. If wall film, local over-rich regions or emissions are to be assessed, the requirements for injector data, droplet models and validation increase.<\/p>\n

Wall wetting and local over-rich mixture<\/h3>\n

Wall film forms when liquid fuel impinges on the piston, cylinder wall, valves or other surfaces. This portion of fuel evaporates later, interacts differently with the flow and can cause local over-rich mixture.
\n\u200d<\/p>\n

On average, the engine may be calibrated correctly while locally unfavourable conditions still arise. This is where unburned hydrocarbons, particulate formation, oil dilution or unstable combustion can become relevant.
\n\u200d<\/p>\n

For development, the injected fuel mass alone is therefore not decisive. What matters is where this mass is actually located at the time of ignition.<\/p>\n

Mixture at the time of ignition<\/h3>\n

At the time of ignition, an ignitable mixture must be present around the ignition source. If the mixture is locally too lean, combustion can become unstable. If it is locally too rich, incomplete combustion, increased emissions or wall-film effects can occur.
\n\u200d<\/p>\n

In high-performance engines, high loads, rapid transients and strong charge motion add further complexity. Tumble, swirl, squish flow and piston-bowl geometry determine how fuel and air are distributed. CFD combustion simulation helps make these local interactions visible.<\/p>\n

What 3D CFD makes visible in e-fuel engine development<\/h2>\n

3D CFD is a spatially resolved field simulation. It shows not only an averaged value, but also local flow, mixing, temperature, pressure, species distribution, wall film and heat transfer.
\n\u200d<\/p>\n

For e-fuel engines, this is relevant because many critical effects arise locally. The test bench provides important measured quantities, but not every local cause can be measured directly. 3D CFD can help understand these causes and sort variants more technically before test bench work.
\n\u200d<\/p>\n

One point remains important: CFD does not provide direct measurements. CFD is a model. Its significance depends on geometry, mesh, boundary conditions, model selection, time step, fuel properties and validation.<\/p>\n

In-cylinder flow and charge motion<\/h3>\n

In-cylinder flow determines how fuel and air come together. Intake port, valve lift, piston motion, piston bowl and combustion chamber geometry generate charge motion such as tumble, swirl and squish flow.
\n\u200d<\/p>\n

These motions influence mixture homogeneity, turbulence at the time of ignition and therefore burn duration, stability and pressure rise. With e-fuels, the effect of this flow can change if spray, evaporation or wall film differ from the reference fuel.
\n\u200d<\/p>\n

3D CFD is particularly valuable here because it does not only show whether an operating point works or does not work. It can show why an operating point becomes critical.<\/p>\n

Temperature and species distribution<\/h3>\n

For combustion and emissions formation, local temperature and species fields are decisive. NOx is strongly dependent on temperature, oxygen availability and residence time. CO and unburned hydrocarbons can be influenced by local mixture, near-wall regions, incomplete reaction or temperatures that are too low.
\n\u200d<\/p>\n

Global metrics show these local differences only to a limited extent. An engine can have a plausible average lambda value and still produce local regions with unfavourable reaction conditions.
\n\u200d<\/p>\n

CFD can show whether temperature peaks, residual-gas fractions, fuel vapour distribution or local air-fuel ratios align with the target quantities.<\/p>\n

Heat transfer and component temperatures<\/h3>\n

E-fuels can also change thermal questions. Heat release, combustion chamber temperature, exhaust gas temperature and wall heat fluxes act on pistons, valves, spark plug, cylinder head and cooling channels.
\n\u200d<\/p>\n

For high-performance engines, thermal behaviour is not a comfort issue. Local temperature peaks can affect component life, material limits, knock tendency, pre-ignition and emissions.
\n\u200d<\/p>\n

If flow, combustion and heat transfer must be assessed together, a coupled view can be useful. Felsaris Engineering<\/a> addresses such questions through simulation-based development, technical system assessment and validation logic.<\/p>\n

Reactive CFD: chemistry, flamelet models and reduced mechanisms<\/h2>\n

If burn rate, heat release and emissions are to be assessed, pure flow simulation is not sufficient. This is where reactive CFD comes in. Reactive CFD couples flow, mixing, heat transfer and chemical reaction.
\n\u200d<\/p>\n

The central question is not which model is the most complex. The central question is which model is robust for the target quantity, operating range and available data.<\/p>\n

Detailed chemistry vs. reduced mechanisms<\/h3>\n

Detailed reaction mechanisms can contain a very large number of species and reactions. This is valuable from a technical perspective, but is often computationally expensive in real full-engine geometries. For this reason, reduced mechanisms are frequently used in engineering development.
\n\u200d<\/p>\n

However, a reduced mechanism must match the target quantity. A mechanism that predicts ignition delay well may still represent NOx or CO inadequately. Conversely, a mechanism can be useful for emission trends without resolving every local flame structure perfectly.
\n\u200d<\/p>\n

The best chemistry is therefore not automatically the largest chemistry. What matters is validation for the target quantity and operating range.<\/p>\n

Target quantities in combustion modelling<\/h3>\n

Typical target quantities in reactive CFD are:
\n\u200d<\/p>\n