
The Danes have given up on their best warships: Lessons for the Royal Navy
Building warships is hard. To those who have been in one this will come as no surprise but for the majority who have not, and who see them like one's phone or car, then this is not well understood. A warship is a complicated assembly of many different complicated things, most of which need to be networked and connected to each other.
Any given class of warship tends to be made in small numbers, too, and there may be major differences even between warships of the same class. The first Arleigh Burke class destroyers in the US Navy, for instance, were 8,400 tons: the latest are 9,700. To some extent, every warship is custom built. And they're so expensive that there isn't usually much scope for prototypes and extensive tests on them before building the actual ship.
The Danes, however, do know all this, having lately found out just how difficult it is to get right. The Danish Chief of Defence recently announced that ongoing attempts to repair and upgrade the Royal Danish Navy's Iver Huitfeldt-class frigates will be abandoned. This is painful for the Danes as these three ships are the most modern in the Danish navy: and, on paper at least, they are the most capable.
The reality, dramatically exposed in the Red Sea last year, is rather different.
In the case of the Iver Huitfeldt herself, an integration failure between different weapons and sensors caused the ship's crew to lose the use of their air defence missiles for up to half an hour during a Houthi attack. They rebooted the command system and subsequently fired more missiles but in the hottest missile environment of recent times, this could have ended very differently. Their fallback option for shooting down drones – twin 76mm cannons – had similar command and control issues with 'up to half' the shells detonating too close to the ship to be effective.
The then Chief of Defence was relieved of his position and the ship removed from theatre: red faces all round (which was better than the alternative).
The problem was that a ship's detection and fire control radars, other sensors, command system, weapons and power delivery train all need to work as one and all at once. The search radar must find the incoming threat, it must then pass the information to the fire-control radar, the fire-control radar needs to get a lock on the threat, the missile system then needs that information – usually a human gets briefly involved here. Then a missile must launch and acquire the threat itself at some point, either using the reflections from the fire-control radar or its own active radar, or both at different times: or it may be guided initially by networked commands, possibly using information provided from a completely different ship, or an aircraft. Meanwhile the ship's installed generation must be powering and cooling all this kit reliably and in many modern cases driving the propulsion too.
Good ships – the US Navy's Arleigh Burke destroyers and the Royal Navy's Type 23 frigates, for example – do all this well. Bad ships – the US Littoral Combat Ships and our early Type 45 destroyers before they were partly fixed – do not.
One of the issues is that all the various systems may not originally have been built to work together. The Arleigh Burkes use Mark 41 missile launch tubes, originally built to work with the Burkes' Aegis combat system and its accompanying, hugely powerful radars: that configuration is proven and reliable.
The Iver Huitfelts use Mark 41 tubes too, but connected to different, European made radars and command systems. In this case of the Red Sea failure, many fingers point at the Thales phased array radar's ability to talk to the command system, but the detail isn't as important as noting the complexity. The problem is obviously a deep-seated one, as it has evidently proved uneconomic to fix.
One lesson that is crystal clear is that no feasible amount of laboratory work, peacetime testing and training can replicate the stresses and strains of combat. Whilst the gun issue was known about before, the missile fire control system problem was not, and only became exposed when real weapons were in the air.
Our Type 45 destroyer operating in the Red Sea, HMS Diamond, had the opposite issue when it shot down a ballistic missile previously thought to be outside its capabilities. This is a nice surprise, but it was a surprise, and it shows again that you don't know until you do it for real. Looking back in time, we had terrible problems in the Falklands and in the Gulf with our old Sea Dart and Sea Wolf missiles. (The sole Sea Dart engagement of 1991, in which an antique Silkworm missile was shot down, was claimed as a success. It should be noted that the Silkworm had already – luckily – missed HMS Gloucester and carried on past before the destroyer managed to shoot it down.)
The difficulty of weapons systems integration is also why I get a little nervous when plug-in or containerised systems are cited as a more flexible way of fighting ships in the future. On paper they can appear to be but if they don't integrate into the command system properly then you quickly end up with the Iver H problem. And by the time you've spent enough money to get around this, then maybe you should have just made that system part of the ship in the first place.
Hashtags

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Telegraph
10 hours ago
- Telegraph
The Danes have given up on their best warships: Lessons for the Royal Navy
Building warships is hard. To those who have been in one this will come as no surprise but for the majority who have not, and who see them like one's phone or car, then this is not well understood. A warship is a complicated assembly of many different complicated things, most of which need to be networked and connected to each other. Any given class of warship tends to be made in small numbers, too, and there may be major differences even between warships of the same class. The first Arleigh Burke class destroyers in the US Navy, for instance, were 8,400 tons: the latest are 9,700. To some extent, every warship is custom built. And they're so expensive that there isn't usually much scope for prototypes and extensive tests on them before building the actual ship. The Danes, however, do know all this, having lately found out just how difficult it is to get right. The Danish Chief of Defence recently announced that ongoing attempts to repair and upgrade the Royal Danish Navy's Iver Huitfeldt-class frigates will be abandoned. This is painful for the Danes as these three ships are the most modern in the Danish navy: and, on paper at least, they are the most capable. The reality, dramatically exposed in the Red Sea last year, is rather different. In the case of the Iver Huitfeldt herself, an integration failure between different weapons and sensors caused the ship's crew to lose the use of their air defence missiles for up to half an hour during a Houthi attack. They rebooted the command system and subsequently fired more missiles but in the hottest missile environment of recent times, this could have ended very differently. Their fallback option for shooting down drones – twin 76mm cannons – had similar command and control issues with 'up to half' the shells detonating too close to the ship to be effective. The then Chief of Defence was relieved of his position and the ship removed from theatre: red faces all round (which was better than the alternative). The problem was that a ship's detection and fire control radars, other sensors, command system, weapons and power delivery train all need to work as one and all at once. The search radar must find the incoming threat, it must then pass the information to the fire-control radar, the fire-control radar needs to get a lock on the threat, the missile system then needs that information – usually a human gets briefly involved here. Then a missile must launch and acquire the threat itself at some point, either using the reflections from the fire-control radar or its own active radar, or both at different times: or it may be guided initially by networked commands, possibly using information provided from a completely different ship, or an aircraft. Meanwhile the ship's installed generation must be powering and cooling all this kit reliably and in many modern cases driving the propulsion too. Good ships – the US Navy's Arleigh Burke destroyers and the Royal Navy's Type 23 frigates, for example – do all this well. Bad ships – the US Littoral Combat Ships and our early Type 45 destroyers before they were partly fixed – do not. One of the issues is that all the various systems may not originally have been built to work together. The Arleigh Burkes use Mark 41 missile launch tubes, originally built to work with the Burkes' Aegis combat system and its accompanying, hugely powerful radars: that configuration is proven and reliable. The Iver Huitfelts use Mark 41 tubes too, but connected to different, European made radars and command systems. In this case of the Red Sea failure, many fingers point at the Thales phased array radar's ability to talk to the command system, but the detail isn't as important as noting the complexity. The problem is obviously a deep-seated one, as it has evidently proved uneconomic to fix. One lesson that is crystal clear is that no feasible amount of laboratory work, peacetime testing and training can replicate the stresses and strains of combat. Whilst the gun issue was known about before, the missile fire control system problem was not, and only became exposed when real weapons were in the air. Our Type 45 destroyer operating in the Red Sea, HMS Diamond, had the opposite issue when it shot down a ballistic missile previously thought to be outside its capabilities. This is a nice surprise, but it was a surprise, and it shows again that you don't know until you do it for real. Looking back in time, we had terrible problems in the Falklands and in the Gulf with our old Sea Dart and Sea Wolf missiles. (The sole Sea Dart engagement of 1991, in which an antique Silkworm missile was shot down, was claimed as a success. It should be noted that the Silkworm had already – luckily – missed HMS Gloucester and carried on past before the destroyer managed to shoot it down.) The difficulty of weapons systems integration is also why I get a little nervous when plug-in or containerised systems are cited as a more flexible way of fighting ships in the future. On paper they can appear to be but if they don't integrate into the command system properly then you quickly end up with the Iver H problem. And by the time you've spent enough money to get around this, then maybe you should have just made that system part of the ship in the first place.


Daily Mirror
a day ago
- Daily Mirror
Oscar winner says Hollywood friends 'cancelled' him as he voted for Donald Trump
Producer Brian Grazer, a long-time Democratic donor, said the reaction he received after admitting he voted for Donald Trump made him feel like he was 'getting cancelled' An Oscar-winning producer has said he felt "cancelled" by his Hollywood friends because he voted for Donald Trump in the 2024 presidential election. Brian Grazer, a long-time Democratic donor who identifies himself as "centrist", opened up about the reaction he received for supporting Trump during a new Fox Nation docuseries titled Art of the Surge. The behind-the-scenes documentary follows Trump's return to the White House - and in one scene, Brian can be seen alongside the then president-elect in a VIP box at the Army-Navy game. On that occasion, Brian book a photo with Trump and confessed to a group of surprised women in the box that he had voted for the Republican. At that point, the women asked him: "You mean, you're not voting for Kamala?" to which he replied: "I just can't do that." Brian explained: "And then, one of them leaned in further, and said, 'Are you voting for Trump?' And I said, 'I am. I swear!'" As part of the series, the producer, known for working on films such as A Beautiful Mind and Apollo 13, confessed the reaction he received made him feel like he was "getting cancelled." The New York Times reported that Brian explained his Trump vote by saying: "As a centrist, it was because I could feel and see Biden's deterioration and the lack of direction in the Democratic Party at that time." The second season of Art of the Surge, produced by former Tucker Carlson Tonight executive producer Justin Wells, is currently streaming on Fox Nation. Before voting for Trump, Brian reportedly donated to Kamala Harris in the past, and also raised money for other Democrats. After his confession emerged, fans took to Reddit to share their reactions - and while some agreed with his thought process, others said the fact that he no longer agreed with the direction of Democrats doesn't justify voting for Republicans, as there are "other choices" too. A person wrote: "Does he understand that if he doesn't like the Democrats he doesn't have to vote for Republicans? There are other choices." Another said: "He knew that as a rich person he'd be safe whichever party won. Not everyone has that privilege." A third expressed: "Every centrist is just a republican that is ashamed to admit it." One asked: "Does centrist mean something totally different in the US? In the UK it's synonymous with liberal (rather than left) and generally describes people who would never vote conservative. But in the US it seems to mean people who could happily vote far right?" Brian isn't the only Hollywood celebrity who voiced support for Trump. Other supporters of the president also include actor Mel Gibson, TV host Dr Phil, comedian Tony Hinchcliffe and actress Victoria Jackson.


BreakingNews.ie
2 days ago
- BreakingNews.ie
Watchdog looking into possibility of aide deletions in Hegseth Signal probe
The Pentagon watchdog is looking into whether any of defence secretary Pete Hegseth's aides was asked to delete Signal messages that may have shared sensitive military information with a reporter, according to two people familiar with the investigation and documents reviewed by The Associated Press. The inspector general's request focuses on how information about the March 15 air strikes on Houthi targets in Yemen was shared on the messaging app. Advertisement It comes as Mr Hegseth is scheduled to testify before Congress next week for the first time since his confirmation hearing. He is likely to face questions under oath not only about his handling of sensitive information but also the wider turmoil at the Pentagon following the departures of several senior aides and an internal investigation over information leaks. Mr Hegseth already has faced questions over the installation of an unsecured internet line in his office that bypassed the Pentagon's security protocols and revelations that he shared details about the military strikes in multiple Signal chats. US President Donald Trump has given his backing to Pete Hegseth (Niall Carson/PA) One of the chats included his wife and brother while the other included President Donald Trump's top national security officials and inadvertently included The Atlantic's editor-in-chief Jeffrey Goldberg. Advertisement Neither the Pentagon nor the inspector general's office immediately responded to Friday requests for comment on the investigation. Besides finding out whether anyone was asked to delete Signal messages, the inspector general also is asking some past and current staffers who were with Mr Hegseth on the day of the strikes who posted the information and who had access to his phone, according to the two people familiar with the investigation and the documents reviewed by the AP. Democratic lawmakers and a small number of Republicans have said that the information Mr Hegseth posted to the Signal chats before the military jets had reached their targets could have put those pilots' lives at risk and that for any lower-ranking members of the military it would have led to their firing. Mr Hegseth has said none of the information was classified. Multiple current and former military officials have said there is no way details with that specificity, especially before a strike took place, would have been OK to share on an unsecured device. Advertisement 'I said repeatedly, nobody is texting war plans,' Mr Hegseth told Fox News Channel in April after reporting emerged about the chat that included his family members. 'I look at war plans every day. What was shared over Signal then and now, however you characterise it, was informal, unclassified co-ordinations, for media co-ordinations and other things. That's what I've said from the beginning.' Mr Trump has made clear that Mr Hegseth continues to have his support, saying during a Memorial Day speech at Arlington National Cemetery in Virginia that the defence secretary 'went through a lot' but 'he's doing really well'. Mr Hegseth has limited his public engagements with the press since the Signal controversy. He has yet to hold a Pentagon press briefing and his spokesman has briefed reporters there only once. Advertisement Signal is a publicly-available app that provides encrypted communications, but it can be hacked and is not approved for carrying classified information. On March 14, one day before the strikes against the Houthis, the defence department cautioned personnel about the vulnerability of the app.